NFP allows the merging of 2 flows together into a single offloaded flow.
In the kernel datapath the packet must match 1 flow, impliment its
actions, recirculate, match the 2nd flow and also impliment its actions.
Merging creates a single flow with all actions from the 2 original flows.

Firmware impliments a tunnel header push as the packet is about to egress
the card. Therefore, if the first merge rule candiate pushes a tunnel,
then the second rule can only have an egress action for a valid merge to
occur (or else the action ordering will be incorrect). This prevents the
pushing of a tunnel header followed by the pushing of a vlan header.

In order to support this behaviour, firmware allows VLAN information to
be encoded in the tunnel push action. If this is non zero then the fw will
push a VLAN after the tunnel header push meaning that 2 such flows with
these actions can be merged (with action order being maintained).

Support tunnel in VLAN pushes by encoding VLAN information in the tunnel
push action of any merge flow requiring this.
Signed-off-by: John Hurley <john.hurley@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Acked-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

TC mirred actions (redirect and mirred) can send to egress or ingress of a
device. Currently only egress is used for hw offload rules.

Modify the intermediate representation for hw offload to include mirred
actions that go to ingress. This gives drivers access to such rules and
can decide whether or not to offload them.
Signed-off-by: John Hurley <john.hurley@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

TC mirred actions can send to egress or ingress on a given netdev. Helpers
exist to detect actions that are mirred to egress. Extend the header file
to include helpers to detect ingress mirred actions.
Signed-off-by: John Hurley <john.hurley@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

TC rules can impliment skbedit actions. Currently actions that modify the
skb mark are passed to offloading drivers via the hardware intermediate
representation in the flow_offload API.

Extend this to include skbedit actions that modify the packet type of the
skb. Such actions may be used to set the ptype to HOST when redirecting a
packet to ingress.
Signed-off-by: John Hurley <john.hurley@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The tc_act header file contains an inline function that checks if an
action is changing the skb mark of a packet and a further function to
extract the mark.

Add similar functions to check for and get skbedit actions that modify
the packet type of the skb.
Signed-off-by: John Hurley <john.hurley@netronome.com>
Reviewed-by: Simon Horman <simon.horman@netronome.com>
Reviewed-by: Jakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Jeff Kirsher says:

====================
100GbE Intel Wired LAN Driver Updates 2019-08-04

This series contains more updates to fm10k from Jake Keller.

Jake removes the unnecessary initialization of some variables to help
resolve static code checker warnings.  Explicitly return success during
resume, since the value of 'err' is always success.  Fixed a issue with
incrementing a void pointer, which can produce undefined behavior.  Used
the __always_unused macro for function templates that are passed as
parameters in functions, but are not used.  Simplified the code by
removing an unnecessary macro in determining the value of NON_Q_VECTORS.
Fixed an issue, using bitwise operations to prevent the low address
overwriting the high portion of the address.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Synchronize PCIe PHY initialization with vendor driver version 8.047.01.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add a helper for MAC OCP read-modify-write operations.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This code piece was inherited from RTL8139 code, the register at
address 0x5c however has a different meaning on RTL8169 and is unused.
So we can remove this.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Dave Taht says:

====================
Two small fq_codel optimizations

These two patches improve fq_codel performance
under extreme network loads. The first patch
more rapidly escalates the codel count under
overload, the second just kills a totally useless
statistic.

(sent together because they'd otherwise conflict)
====================
Signed-off-by: Dave Taht <dave.taht@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

It is almost impossible to get anything other than a 0 out of
flow->dropped statistic with a tc class dump, as it resets to 0
on every round.

It also conflates ecn marks with drops.

It would have been useful had it kept a cumulative drop count, but
it doesn't. This patch doesn't change the API, it just stops
tracking a stat and state that is impossible to measure and nobody
uses.
Signed-off-by: Dave Taht <dave.taht@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

In the field fq_codel is often used with a smaller memory or
packet limit than the default, and when the bulk dropper is hit,
the drop pattern bifircates into one that more slowly increases
the codel drop rate and hits the bulk dropper more than it should.

The scan through the 1024 queues happens more often than it needs to.

This patch increases the codel count in the bulk dropper, but
does not change the drop rate there, relying on the next codel round
to deliver the next packet at the original drop rate
(after that burst of loss), then escalate to a higher signaling rate.
Signed-off-by: Dave Taht <dave.taht@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

After commit 171a9bae ("staging/octeon: Allow test build on
!MIPS"), the following combination of configs cause a few Kconfig
warnings and build errors (distilled from arm allyesconfig and Randy's
randconfig builds):

    CONFIG_NETDEVICES=y
    CONFIG_STAGING=y
    CONFIG_COMPILE_TEST=y

and CONFIG_OCTEON_ETHERNET as either a module or built-in.

WARNING: unmet direct dependencies detected for MDIO_OCTEON
  Depends on [n]: NETDEVICES [=y] && MDIO_DEVICE [=y] && MDIO_BUS [=y]
&& 64BIT [=n] && HAS_IOMEM [=y] && OF_MDIO [=n]
  Selected by [y]:
  - OCTEON_ETHERNET [=y] && STAGING [=y] && (CAVIUM_OCTEON_SOC ||
COMPILE_TEST [=y]) && NETDEVICES [=y]

In file included from ../drivers/net/phy/mdio-octeon.c:14:
../drivers/net/phy/mdio-cavium.h:111:36: error: implicit declaration of
function ‘writeq’; did you mean ‘writel’?
[-Werror=implicit-function-declaration]
  111 | #define oct_mdio_writeq(val, addr) writeq(val, (void *)addr)
      |                                    ^~~~~~

CONFIG_64BIT is not strictly necessary if the proper readq/writeq
definitions are included from io-64-nonatomic-lo-hi.h.

CONFIG_OF_MDIO is not needed when CONFIG_COMPILE_TEST is enabled because
of commit f9dc9ac5 ("of/mdio: Add dummy functions in of_mdio.h.").

Fixes: 171a9bae ("staging/octeon: Allow test build on !MIPS")
Reported-by: kbuild test robot <lkp@intel.com>
Reported-by: Mark Brown <broonie@kernel.org>
Reported-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Nathan Chancellor <natechancellor@gmail.com>
Acked-by: Randy Dunlap <rdunlap@infradead.org> # build-tested
Signed-off-by: David S. Miller <davem@davemloft.net>

Merge the CPU port registers dump into the master interface registers
dump through ethtool, by nesting the ethtool_drvinfo and ethtool_regs
structures of the CPU port into the dump.

drvinfo->regdump_len will contain the full data length, while regs->len
will contain only the master interface registers dump length.

This allows for example to dump the CPU port registers on a ZII Dev
C board like this:

    # ethtool -d eth1
    0x004:                                              0x00000000
    0x008:                                              0x0a8000aa
    0x010:                                              0x01000000
    0x014:                                              0x00000000
    0x024:                                              0xf0000102
    0x040:                                              0x6d82c800
    0x044:                                              0x00000020
    0x064:                                              0x40000000
    0x084: RCR (Receive Control Register)               0x47c00104
        MAX_FL (Maximum frame length)                   1984
        FCE (Flow control enable)                       0
        BC_REJ (Broadcast frame reject)                 0
        PROM (Promiscuous mode)                         0
        DRT (Disable receive on transmit)               0
        LOOP (Internal loopback)                        0
    0x0c4: TCR (Transmit Control Register)              0x00000004
        RFC_PAUSE (Receive frame control pause)         0
        TFC_PAUSE (Transmit frame control pause)        0
        FDEN (Full duplex enable)                       1
        HBC (Heartbeat control)                         0
        GTS (Graceful transmit stop)                    0
    0x0e4:                                              0x76735d6d
    0x0e8:                                              0x7e9e8808
    0x0ec:                                              0x00010000
    .
    .
    .
    88E6352  Switch Port Registers
    ------------------------------
    00: Port Status                            0x4d04
          Pause Enabled                        0
          My Pause                             1
          802.3 PHY Detected                   0
          Link Status                          Up
          Duplex                               Full
          Speed                                100 or 200 Mbps
          EEE Enabled                          0
          Transmitter Paused                   0
          Flow Control                         0
          Config Mode                          0x4
    01: Physical Control                       0x003d
          RGMII Receive Timing Control         Default
          RGMII Transmit Timing Control        Default
          200 BASE Mode                        100
          Flow Control's Forced value          0
          Force Flow Control                   0
          Link's Forced value                  Up
          Force Link                           1
          Duplex's Forced value                Full
          Force Duplex                         1
          Force Speed                          100 or 200 Mbps
    .
    .
    .
Signed-off-by: Vivien Didelot <vivien.didelot@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Ido Schimmel says:

====================
drop_monitor: Various improvements and cleanups

This patchset performs various improvements and cleanups in drop monitor
with no functional changes intended. There are no changes in these
patches relative to the RFC I sent two weeks ago [1].

A followup patchset will extend drop monitor with a packet alert mode in
which the dropped packet is notified to user space instead of just a
summary of recent drops. Subsequent patchsets will add the ability to
monitor hardware originated drops via drop monitor.

[1] https://patchwork.ozlabs.org/cover/1135226/
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Each operation from user space should be protected by the global drop
monitor mutex. Use the pre_doit / post_doit hooks to take / release the
lock instead of doing it explicitly in each function.
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add various extack messages to make drop_monitor more user friendly.
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Remove multiple blank lines which are visually annoying and useless.

This suppresses the "Please don't use multiple blank lines" checkpatch
messages.
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

While 'per_cpu_dm_data' is a per-CPU variable, its 'skb' and
'send_timer' fields can be accessed concurrently by the CPU sending the
netlink notification to user space from the workqueue and the CPU
tracing kfree_skb(). This spinlock is meant to protect against that.

Document its scope and suppress the checkpatch message "spinlock_t
definition without comment".
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The 'trace_state_mutex' does not only protect the global 'trace_state'
variable, but also the global 'hw_stats_list'.

Subsequent patches are going add more operations from user space to
drop_monitor and these all need to be mutually exclusive.

Rename 'trace_state_mutex' to the more fitting 'net_dm_mutex' name and
document its scope.
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The error code 'ENOTSUPP' is reserved for use with NFS. Use 'EOPNOTSUPP'
instead.
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This Kconfig option is unused, drop it.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Andrew Lunn <andrew@lunn.ch>
Cc: David S. Miller <davem@davemloft.net>
Cc: Florian Fainelli <f.fainelli@gmail.com>
Cc: Tristram Ha <Tristram.Ha@microchip.com>
Cc: Vivien Didelot <vivien.didelot@gmail.com>
Cc: Woojung Huh <woojung.huh@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Merge the two headers into one, no functional change.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Andrew Lunn <andrew@lunn.ch>
Cc: David S. Miller <davem@davemloft.net>
Cc: Florian Fainelli <f.fainelli@gmail.com>
Cc: Tristram Ha <Tristram.Ha@microchip.com>
Cc: Vivien Didelot <vivien.didelot@gmail.com>
Cc: Woojung Huh <woojung.huh@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Remove ksz_port_cleanup(), which is unused. Add missing include
"ksz_common.h", which fixes the following warning when built with
make ... W=1

drivers/net/dsa/microchip/ksz_common.c:23:6: warning: no previous prototype for ‘...’ [-Wmissing-prototypes]

Note that the order of the headers cannot be swapped, as that would
trigger missing forward declaration errors, which would indicate the
way forward is to merge the two headers into one.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Andrew Lunn <andrew@lunn.ch>
Cc: David S. Miller <davem@davemloft.net>
Cc: Florian Fainelli <f.fainelli@gmail.com>
Cc: Tristram Ha <Tristram.Ha@microchip.com>
Cc: Vivien Didelot <vivien.didelot@gmail.com>
Cc: Woojung Huh <woojung.huh@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The driver is relying on zero'ed allocated memory and does not
explicitly call atomic_set() to initialize the ref counts to 0.  Add
these atomic_set() calls so that it will be more straight forward
to convert atomic ref counts to refcount_t.
Reported-by: Chuhong Yuan <hslester96@gmail.com>
Cc: Rasesh Mody <rmody@marvell.com>
Cc: <GR-Linux-NIC-Dev@marvell.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Simplify the unlock path in __ip6_rt_update_pmtu by using a
single point where rcu_read_unlock is called.
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Saeed Mahameed says:

====================
mlx5-updates-2019-08-01

Misc updates for mlx5 netdev driver:

1) Ingress rate support for E-Switch vports from Eli.
2) Gavi introduces flow counters bulk allocation and pool,
   To improve the performance of flow counter acquisition.
3) From Tariq, micro improvements for tx path
4) From Shay, small improvement for XDP TX MPWQE inline flow.
5) Aya provides some cleanups for tx devlink health reporters.
6) Saeed, refactor checksum handling into a single function.
7) Tonghao, allows dropping specific tunnel packets.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

There is a spelling mistake in an error messgae. Fix it.
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Acked-by: Shuah Khan <skhan@linuxfoundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Fix assignment of the FM10K_FAULT_ADDR_LO register into fault->address
by using a bit-wise |= operation. Without this, the low address is
completely overwriting the high potion of the address. This caused the
fault to incorrectly return only the lower 32 bits of the fault address.

This issue was detected by cppcheck and resolves the following warnings
produced by that tool:

[fm10k_pf.c:1668] -> [fm10k_pf.c:1670]: (style) Variable
'fault->address' is reassigned a value before the old one has been used.

[fm10k_pf.c:1669] -> [fm10k_pf.c:1670]: (style) Variable
'fault->address' is reassigned a value before the old one has been used.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

The driver currently uses a macro to decide whether we should use
NON_Q_VECTORS_PF or NON_Q_VECTORS_VF.

However, we also define NON_Q_VECTORS_VF to the same value as
NON_Q_VECTORS_PF. This means that the macro NON_Q_VECTORS(hw) will
always return the same value.

Let's just remove this macro, and replace it directly with an enum value
on the enum non_q_vectors.

This was detected by cppcheck and fixes the following warnings when
building with BUILD=KERNEL

[fm10k_ethtool.c:1123]: (style) Same value in both branches of ternary
operator.

[fm10k_ethtool.c:1142]: (style) Same value in both branches of ternary
operator.

[fm10k_main.c:1826]: (style) Same value in both branches of ternary
operator.

[fm10k_main.c:1849]: (style) Same value in both branches of ternary
operator.

[fm10k_main.c:1858]: (style) Same value in both branches of ternary
operator.

[fm10k_pci.c:901]: (style) Same value in both branches of ternary
operator.

[fm10k_pci.c:1040]: (style) Same value in both branches of ternary
operator.

[fm10k_pci.c:1726]: (style) Same value in both branches of ternary
operator.

[fm10k_pci.c:1763]: (style) Same value in both branches of ternary
operator.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

Several functions in the fm10k driver have specific function templates,
as they are used as function pointers. The parameters in these functions
are not always used. Explicitly mark unused parameters with the
__always_unused macro, so that the compiler will not warn about them
when building with the -Wunused-parameter warning enabled.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

The page_addr variable is a void pointer. Incrementing it before calling
prefetch is technically undefined. Fix this by casting it to a u8*
pointer before incrementing it. This ensures that we increment the
pointer value in byte units, instead of relying on this undefined
behavior.

This was detected by cppcheck, and resolves the following warning
produced by that tool:

[fm10k_main.c:328]: (portability) 'page_addr' is of type 'void *'. When
using void pointers in calculations, the behaviour is undefined.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

In the fm10k_handle_resume function, return 0 explicitly at the end of
the function instead of returning the err value.

This was detected by cppcheck and resolves the following style warning
produced by that tool:

[fm10k_pci.c:2768] -> [fm10k_pci.c:2787]: (warning) Identical condition
'err', second condition is always false
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

The local variable 'size' in fm10k_dfwd_add_station is initialized, but
is always re-assigned immediately before use. Remove this unnecessary
initialization.

This was detected by cppcheck and resolves the following warning
produced by that tool:

[fm10k_netdev.c:1466]: (style) Variable 'size' is assigned a value that is never used.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

The local variable err in several functions in the fm10k_netdev.c file
is initialized with a value that is never used. The err value is
immediately re-assigned in all cases where it will be checked. Remove
the unnecessary initializers.

This was detected by cppcheck and resolves the following warnings
produced by that tool:

[fm10k_netdev.c:999] -> [fm10k_netdev.c:1004]: (style) Variable 'err' is
reassigned a value before the old one has been used.

[fm10k_netdev.c:1019] -> [fm10k_netdev.c:1024]: (style) Variable 'err'
is reassigned a value before the old one has been used.

[fm10k_netdev.c:64]: (style) Variable 'err' is assigned a value that is
never used.

[fm10k_netdev.c:131]: (style) Variable 'err' is assigned a value that
is never used.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

The err variable in the fm10k_tlv_attr_parse function is initialized
with zero. However, the function never reads err without first assigning
it from a function call. Remove this unnecessary initialization.

This was detected by cppcheck and resolves the following warning
produced by that tool:

[fm10k_tlv.c:498]: (style) Variable 'err' is assigned a value that is
never used.
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Andrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>

David Ahern says:

====================
net: Add functional tests for L3 and L4

This is a port the functional test cases created during the development
of the VRF feature. It covers various permutations of icmp, tcp and udp
for IPv4 and IPv6 including negative tests.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Add use case section to fcnal-test.

Initial test is VRF based with a bridge and vlans. The commands
stem from bug reports fixed by:

a173f066 ("netfilter: bridge: Don't sabotage nf_hook calls from an l3mdev")
cd642898 ("netfilter: bridge: Don't sabotage nf_hook calls for an l3mdev slave")
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add IPv6 netfilter tests to send tcp reset or icmp unreachable for a
port. Initial tests are VRF only.
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add netfilter tests to send tcp reset or icmp unreachable for a port.
Initial tests are VRF only.
Signed-off-by: David Ahern <dsahern@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>