app/testpmd: add eCPRI in flow creation patterns

In order to verify offloading of eCPRI protocol via flow rules, the
command line of flow creation should support the parsing of the eCPRI
pattern.

Based on the specification, one eCPRI message will have the common
header and payload. Payload format is various based on the type field
of the common header. Fixed strings will be used instead of integer
to make the CLI easy for auto-completion.

The testpmd command line examples of flow to match eCPRI item are
listed below:
  1. flow create 0 ... pattern eth / ecpri / end actions ...
    This is to match all eCPRI messages.
  2. flow create 0 ... pattern eth / ecpri common type rtc_ctrl / end actions ...
    This is to match all eCPRI messages with the type #2 - "Real-Time
    Control Data".
  3. flow create 0 ... pattern eth / ecpri common type iq_data pc_id is [U16Int] / end actions ...
    This is to match eCPRI messages with the type #0 - "IQ Data", and
    the physical channel ID 'pc_id' of the messages is a specific
    value. Since the sequence ID is changeable, there is no need to
    match that field in the flow.
Currently, only type #0, #2 and #5 will be supported.

Since eCPRI could be over Ethernet layer (or after .1Q) and UDP
layer, it is the PMD driver's responsibility to check whether eCPRI
is supported and which protocol stack is supported. Network byte
order should be used for eCPRI header, the same as other headers.

Signed-off-by: Bing Zhao <bingz@mellanox.com>
Acked-by: Ori Kam <orika@mellanox.com>

ethdev: add eCPRI key fields to flow API

Add a new item "rte_flow_item_ecpri" in order to match eCRPI header.

eCPRI is a packet based protocol used in the fronthaul interface of
5G networks. Header format definition could be found in the
specification via the link below:
https://www.gigalight.com/downloads/standards/ecpri-specification.pdf

eCPRI message can be over Ethernet layer (.1Q supported also) or over
UDP layer. Message header formats are the same in these two variants.

Signed-off-by: Bing Zhao <bingz@mellanox.com>
Acked-by: Ori Kam <orika@mellanox.com>
Acked-by: Olivier Matz <olivier.matz@6wind.com>

net/bnxt: fix freeing filters on flow creation failure

This patch does following things:
1. Added a wrapper function bnxt_clear_one_vnic_filter()
   for destroying the filters in hw. This will avoid duplicate
   code in many places.
2. When flow create fails due to an already existing mark id
   for the new flow id created, fixed to destroy the hw
   filter created.
3. Re-arranged code to move a log and list update to right place.

Fixes: 9db66782bd06 ("net/bnxt: fix supporting zero mark ID with RSS action")
Fixes: 5ef3b79fdfe6 ("net/bnxt: support flow filter ops")
Cc: stable@dpdk.org
Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/bnxt: fix flow error on filter creation

If set_em_filter/set_ntuple_filter cmds fails for some reason,
driver is not filling the "rte_flow_error" string buffer.
Same is the case when flow create fails due to an already
existing mark id for the new flow id created.
This leads to a crash in testpmd while trying to print the
error message.

Fixes: 5c1171c97216 ("net/bnxt: refactor filter/flow")
Fixes: 9db66782bd06 ("net/bnxt: fix supporting zero mark ID with RSS action")
Cc: stable@dpdk.org
Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/bnxt: fix unnecessary HWRM command

During probe, driver issues HWRM_CFA_ADV_FLOW_MGNT_QCAPS command.
But it is not using the command response anywhere which makes the
fw call redundant.

Remove the unnecessary HWRM_CFA_ADV_FLOW_MGNT_QCAPS call to fw.
Remove the redundant flow_flags in bnxt struct.

Fixes: afef822b2e1b ("net/bnxt: support creating SMAC and inner DMAC filters")
Cc: stable@dpdk.org
Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/bnxt: remove unused enum declaration

"enum bnxt_hw_context" is defined in the header file, but is not
used anywhere.

Fixes: 9738793f28ec ("net/bnxt: add VNIC functions and structs")
Cc: stable@dpdk.org
Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/bnxt: fix unnecessary memory allocation

VFs are not privileged to issue HWRM_PORT_LED_QCFG/CFG.
There is no need to allocate "bp->leds" memory.

Fixes: 205b74295282 ("net/bnxt: fix allocation of LED config info")
Cc: stable@dpdk.org
Signed-off-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/bonding: fix LACP negotiation

When two host is connected directly without any devices like switch,
rx_machine_update would receiving partner LACP negotiation packets,
and partner's port mac is filled with zeros in this packet, which is
different with internal's mode4 mac. So in this situation, it would
never go rx_machine branch and then execute mac swap for negotiation!
Thus bond mode 4 will negotiation failed.

Fixes: 56cbc0817399 ("net/bonding: fix LACP negotiation")
Cc: stable@dpdk.org
Signed-off-by: Yicai Lu <luyicai@huawei.com>
Reviewed-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>

app/testpmd: add 5-tuple swap forwarding engine

The new 5-tuple swap engine swaps:
source and destination mac address,
source and destination address in ipv4/ipv6,
source and destination port in UDP/TCP.

The forwarding engine will parse each layer
and swap it, and will stop when the next
layer doesn't match.

The mentioned headers of ICMP/ARP/Multicast
packets will be swapped as well according to
matching layers.

usage: --forward-mode=5tswap

Signed-off-by: Shiri Kuzin <shirik@mellanox.com>
Acked-by: Matan Azrad <matan@mellanox.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

drivers/net: delete HW rings while freeing queues

Delete memzones for HW rings in igb and ixgbe while freeing queues

Updated igb, ixgbe, i40e, ice & em drivers.

Signed-off-by: Renata Saiakhova <renata.saiakhova@ekinops.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

ethdev: add function to release HW rings

Free previously allocated memzone for HW rings

Signed-off-by: Renata Saiakhova <renata.saiakhova@ekinops.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Andrew Rybchenko <arybchenko@solarflare.com>

drivers/net: fix exposing internal headers

Using '__rte_internal' tag in 'rte_ethdev_driver.h' causing build error
for applications and examples. Because they don't define
'ALLOW_INTERNAL_API' flag and '__rte_internal' causes the error.
This patch is preparation for future '__rte_internal' usage.

At first place, applications/examples should not include
'rte_ethdev_driver.h', this is happening because of PMD public header
files include 'rte_ethdev_driver.h' by mistake.

Updated PMD public header files to not include internal header files.

But for unit test application, 'app/test', enable accessing internal
APIs, since some unit tests need them.

Fixes: ffc905f3b856 ("ethdev: separate driver APIs")
Fixes: ec0dec44ecb9 ("net/atlantic: enable MACsec configuration")
Cc: stable@dpdk.org
Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Thomas Monjalon <thomas@monjalon.net>

net/nfp: fix RSS hash configuration reporting

Prior to this fix the NFP PMD implementation of the .rss_hash_conf_get
callback did not propagate the current hardware state of rss_hf back up
to the caller. Users of the hash_conf_get callback would receive an
incorrect representation of what the RSS configuration currently is in
hardware.

Fixes: 934e4c60fbff ("nfp: add RSS")
Cc: stable@dpdk.org
Signed-off-by: Heinrich Kuhn <heinrich.kuhn@netronome.com>
Signed-off-by: Simon Horman <simon.horman@netronome.com>

app/testpmd: add Tx scheduling command

This commit adds testpmd capability to provide timestamps on the packets
being sent in the txonly mode. This includes:

 - SEND_ON_TIMESTAMP support
   new device Tx offload capability support added, example:

     testpmd> port config 0 tx_offload send_on_timestamp on

 - set txtimes, registers field and flag, example:

     testpmd> set txtimes 1000000,0

   This command enables the packet send scheduling on timestamps if
   the first parameter is not zero, generic format:

     testpmd> set txtimes (inter),(intra)

   where:

     inter - is the delay between the bursts in the device clock units.
     If "intra" (next parameter) is zero, this is the time between the
     beginnings of the first packets in the neighbour bursts, if "intra"
     is not zero, "inter" specifies the time between the beginning of
     the first packet of the current burst and the beginning of the last
     packet of the previous burst. If "inter"parameter is zero the send
     scheduling on timestamps is disabled (default).

     intra - is the delay between the packets within the burst specified
     in the device clock units. The number of packets in the burst is
     defined by regular burst setting. If "intra" parameter is zero no
     timestamps provided in the packets excepting  the first one in the
     burst.

     As the result the bursts of packet will be transmitted with
     specific delay between the packets within the burst and specific
     delay between the bursts. The rte_eth_read_clock() is supposed to
     be engaged to get the current device clock value and provide the
     reference for the timestamps. If there is no supported
     rte_eth_read_clock() there will be no provided send scheduling on
     the device.

 - show txtimes, displays the timing settings
 - txonly burst time pattern

Signed-off-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

mbuf: introduce accurate packet Tx scheduling

There is the requirement on some networks for precise traffic timing
management. The ability to send (and, generally speaking, receive)
the packets at the very precisely specified moment of time provides
the opportunity to support the connections with Time Division
Multiplexing using the contemporary general purpose NIC without involving
an auxiliary hardware. For example, the supporting of O-RAN Fronthaul
interface is one of the promising features for potentially usage of the
precise time management for the egress packets.

The main objective of this patchset is to specify the way how applications
can provide the moment of time at what the packet transmission must be
started and to describe in preliminary the supporting this feature
from mlx5 PMD side [1].

The new dynamic timestamp field is proposed, it provides some timing
information, the units and time references (initial phase) are not
explicitly defined but are maintained always the same for a given port.
Some devices allow to query rte_eth_read_clock() that will return
the current device timestamp. The dynamic timestamp flag tells whether
the field contains actual timestamp value. For the packets being sent
this value can be used by PMD to schedule packet sending.

The device clock is opaque entity, the units and frequency are
vendor specific and might depend on hardware capabilities and
configurations. If might (or not) be synchronized with real time
via PTP, might (or not) be synchronous with CPU clock (for example
if NIC and CPU share the same clock source there might be no
any drift between the NIC and CPU clocks), etc.

After PKT_RX_TIMESTAMP flag and fixed timestamp field supposed
deprecation and obsoleting, these dynamic flag and field might be
used to manage the timestamps on receiving datapath as well. Having
the dedicated flags for Rx/Tx timestamps allows applications not
to perform explicit flags reset on forwarding and not to promote
received timestamps to the transmitting datapath by default.
The static PKT_RX_TIMESTAMP is considered as candidate to become
the dynamic flag and this move should be discussed.

When PMD sees the "rte_dynfield_timestamp" set on the packet being sent
it tries to synchronize the time of packet appearing on the wire with
the specified packet timestamp. If the specified one is in the past it
should be ignored, if one is in the distant future it should be capped
with some reasonable value (in range of seconds). These specific cases
("too late" and "distant future") can be optionally reported via
device xstats to assist applications to detect the time-related
problems.

There is no any packet reordering according timestamps is supposed,
neither within packet burst, nor between packets, it is an entirely
application responsibility to generate packets and its timestamps
in desired order. The timestamps can be put only in the first packet
in the burst providing the entire burst scheduling.

PMD reports the ability to synchronize packet sending on timestamp
with new offload flag:

This is palliative and might be replaced with new eth_dev API
about reporting/managing the supported dynamic flags and its related
features. This API would break ABI compatibility and can't be introduced
at the moment, so is postponed to 20.11.

For testing purposes it is proposed to update testpmd "txonly"
forwarding mode routine. With this update testpmd application generates
the packets and sets the dynamic timestamps according to specified time
pattern if it sees the "rte_dynfield_timestamp" is registered.

The new testpmd command is proposed to configure sending pattern:

set tx_times <burst_gap>,<intra_gap>

<intra_gap> - the delay between the packets within the burst
              specified in the device clock units. The number
              of packets in the burst is defined by txburst parameter

<burst_gap> - the delay between the bursts in the device clock units

As the result the bursts of packet will be transmitted with specific
delays between the packets within the burst and specific delay between
the bursts. The rte_eth_read_clock is supposed to be engaged to get the
current device clock value and provide the reference for the timestamps.

[1] http://patches.dpdk.org/patch/73714/

Signed-off-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
Acked-by: Olivier Matz <olivier.matz@6wind.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net: use named constants for deprecated QinQ TPIDs

Add named constants for deprecated QinQ TPIDs.
Update drivers which have already been using existing
TPID named constants from librte_net to use the
new named constants rather than magic numbers.

Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

app/testpmd: fix CPU cycles per packet stats on Tx modes

In txonly and flowgen forwarding mode, calculating CPU per packets with
total received packets is not accurate. Use total transmitted packets
for these cases.

The error output under txonly mode:
testpmd> show fwd stats all

---------------------- Forward statistics for port 0  -------------------
RX-packets: 0              RX-dropped: 0             RX-total: 0
TX-packets: 3582891927     TX-dropped: 401965824     TX-total: 3984857751
TX-bursts : 86381636 [0% of 0 pkts + 85% of 64 pkts + 15% of 32 pkts]
-------------------------------------------------------------------------

---------------------- Forward statistics for port 1  -------------------
RX-packets: 1              RX-dropped: 394351696     RX-total: 394351697
TX-packets: 3582890632     TX-dropped: 401965568     TX-total: 3984856200
TX-bursts : 86381679 [0% of 0 pkts + 85% of 64 pkts + 15% of 32 pkts]
-------------------------------------------------------------------------

+++++++++++++++ Accumulated forward statistics for all ports+++++++++++++
RX-packets: 1              RX-dropped: 394351696     RX-total: 394351697
TX-packets: 7165782559     TX-dropped: 803931392     TX-total: 7969713951
+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

CPU cycles/packet=54984156291.00 \
(total cycles=54984156291 / total RX packets=1) at 200 MHz Clock

Fixes: 53324971a14e ("app/testpmd: display/clear forwarding stats on demand")
Cc: stable@dpdk.org
Signed-off-by: Phil Yang <phil.yang@arm.com>
Reviewed-by: Ruifeng Wang <ruifeng.wang@arm.com>
Reviewed-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
Tested-by: Ferruh Yigit <ferruh.yigit@intel.com>

app/testpmd: remove softnic forward mode

Softnic can be used like other virtual devices without
needing any special mode. Therefore, remove softnic mode
from testpmd app. Documentation is updated as well.

Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>
Acked-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>

doc: announce marking ethdev internal symbol

The APIs are marked in the doxygen comment but better to mark the
symbols too. This is planned for v20.11 release.

Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Andrew Rybchenko <arybchenko@solarflare.com>
Acked-by: Ray Kinsella <mdr@ashroe.eu>
Acked-by: Thomas Monjalon <thomas@monjalon.net>

net/enic: support burst mode info

Add Rx/Tx burst mode getter handlers.

Signed-off-by: Hyong Youb Kim <hyonkim@cisco.com>
Reviewed-by: John Daley <johndale@cisco.com>

net/ice: support RSS for IPv6 prefix

RSS for IPv6 prefix fields are supported in this patch, so that we
can use prefixes instead of full IPv6 address for RSS. These prefixes
include the first 32, 48, 64 bits of both SRC and DST IPv6 address.

Signed-off-by: Junfeng Guo <junfeng.guo@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>

app/testpmd: support extended RSS offload types

This patch adds testpmd cmdline support for IPv6 prefix.

Signed-off-by: Junfeng Guo <junfeng.guo@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

ethdev: add RSS types for IPv6 prefix

This patch defines new RSS offload types for IPv6 prefix with 32, 40,
48, 56, 64, 96 bits of both SRC and DST IPv6 address.
Ref https://tools.ietf.org/html/rfc6052.

Signed-off-by: Junfeng Guo <junfeng.guo@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/hinic/base: convert error value to ETIMEDOUT

Following commit updated the error codes:
commit 2ae8e130cf21 ("net/hinic/base: modify returned error values")

In that commit 'ETIME' errors are not used because it is not supported
by FreeBSD, instead in this patch converting relevant error codes to
'ETIMEDOUT'.

Signed-off-by: Xiaoyun Wang <cloud.wangxiaoyun@huawei.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

ethdev: fix VLAN offloads set if no relative capabilities

Currently, there is a potential problem that calling the API function
rte_eth_dev_set_vlan_offload to start VLAN hardware offloads which the
driver does not support. If the PMD driver does not support certain VLAN
hardware offloads and does not check for it, the hardware setting will
not change, but the VLAN offloads in dev->data->dev_conf.rxmode.offloads
will be turned on.

It is supposed to check the hardware capabilities to decide whether the
relative callback needs to be called just like the behavior in the API
function named rte_eth_dev_configure. And it is also needed to cleanup
duplicated checks which are done in some PMDs. Also, note that it is
behaviour change for some PMDs which simply ignore (with error/warning
log message) unsupported VLAN offloads, but now it will fail.

Fixes: a4996bd89c42 ("ethdev: new Rx/Tx offloads API")
Fixes: 0ebce6129bc6 ("net/dpaa2: support new ethdev offload APIs")
Fixes: f9416bbafd98 ("net/enic: remove VLAN filter handler")
Fixes: 4f7d9e383e5c ("fm10k: update vlan offload features")
Fixes: fdba3bf15c7b ("net/hinic: add VLAN filter and offload")
Fixes: b96fb2f0d22b ("net/i40e: handle QinQ strip")
Fixes: d4a27a3b092a ("nfp: add basic features")
Fixes: 56139e85abec ("net/octeontx: support VLAN filter offload")
Fixes: ba1b3b081edf ("net/octeontx2: support VLAN offloads")
Fixes: d87246a43759 ("net/qede: enable and disable VLAN filtering")
Cc: stable@dpdk.org
Signed-off-by: Chengchang Tang <tangchengchang@huawei.com>
Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Acked-by: Andrew Rybchenko <arybchenko@solarflare.com>
Acked-by: Hyong Youb Kim <hyonkim@cisco.com>
Acked-by: Sachin Saxena <sachin.saxena@nxp.com>
Acked-by: Xiaoyun Wang <cloud.wangxiaoyun@huawei.com>
Acked-by: Harman Kalra <hkalra@marvell.com>
Acked-by: Jeff Guo <jia.guo@intel.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>

ethdev: fix data room size verification in Rx queue setup

In the rte_eth_rx_queue_setup API function, the local variable named
mbp_buf_size, which is the data room size of the input parameter mp,
is checked to guarantee that each memory chunk used for net device
in the mbuf is bigger than the min_rx_bufsize. But if mbp_buf_size is
less than RTE_PKTMBUF_HEADROOM, the value of the following  statement
will be a large number since the mbp_buf_size is a unsigned value.
    mbp_buf_size - RTE_PKTMBUF_HEADROOM
As a result, it will cause a segment fault in this situation.

This patch fixes it by modify the check condition to guarantee that the
local variable named mbp_buf_size is bigger than RTE_PKTMBUF_HEADROOM.

Fixes: af75078fece3 ("first public release")
Cc: stable@dpdk.org
Signed-off-by: Chengchang Tang <tangchengchang@huawei.com>
Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Reviewed-by: Andrew Rybchenko <arybchenko@solarflare.com>
Acked-by: Sachin Saxena <sachin.saxena@nxp.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>

net/dpaa2: avoid multiple IOVA conversions

Avoids recalculation of DPAA2_IOVA_TO_VADDR, since this is somewhat
costly function on no-IOMMU setups.

Brings about 0.5 Gbps improvement at 19 Gbps speeds on LX2160 with 512b
packets.

Signed-off-by: Roman Kapl <rka@sysgo.com>
Reviewed-by: Hemant Agrawal <hemant.agrawal@nxp.com>

net/enic: add more flow manager actions to dump

The following commit introduced several new actions. Make the dump
function to print those actions.

commit 6faf81f1d2c3 ("net/enic: update flow manager API")

Signed-off-by: Hyong Youb Kim <hyonkim@cisco.com>
Reviewed-by: John Daley <johndale@cisco.com>

net/enic: support VLAN push and pop flow actions

Flow manager API includes push/pop actions, so support corresponding
DPDK flow actions.

Signed-off-by: Hyong Youb Kim <hyonkim@cisco.com>
Reviewed-by: John Daley <johndale@cisco.com>

common/mlx5: fix physical port name pattern recognition

This patch makes the Infiniband device physical port name
recognition more strict. Currently mlx5 PMD might recognize
the names like "pf0sf0" erroneously as "pf0" and the wrong
device type (host PF representor) is reported.

The names like "pf0sf0" belong to PCI subfunctions which
is currently not supported by mlx5 PMD and this false
recognition must be eliminated.

Fixes: 420bbdae89f2 ("net/mlx5: fix host physical function representor naming")
Cc: stable@dpdk.org
Signed-off-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>

net/mlx5: add ConnectX-6 Lx device ID

This adds the ConnectX-6 Lx device id to the list of supported
Mellanox devices that run the MLX5 PMD.
The device is still in development stage.

Signed-off-by: Ali Alnubani <alialnu@mellanox.com>
Acked-by: Raslan Darawsheh <rasland@mellanox.com>

doc: fix a typo in mlx5 guide

Fixes: ecb160456aed ("net/mlx5: add device parameter for MPRQ stride size")
Cc: stable@dpdk.org
Signed-off-by: Ali Alnubani <alialnu@mellanox.com>
Acked-by: Raslan Darawsheh <rasland@mellanox.com>

net/iavf: support RSS for IPv4 IPv6 mix of GTP

RSS for GTP with outer & inner ipv4 & ipv6 combination are supported
in this patch, so that we can process RSS based on inner 5 tuples or
3 tuples of all the cases below of GTP packets:
        1. ipv4 (outer) + ipv4 (inner)
        2. ipv4 (outer) + ipv6 (inner)
        3. ipv6 (outer) + ipv4 (inner)
        4. ipv6 (outer) + ipv6 (inner)

Signed-off-by: Junfeng Guo <junfeng.guo@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>

net/ice: support RSS for IPv4 IPv6 mix of GTP

RSS for GTP with outer & inner ipv4 & ipv6 combination are supported
in this patch, so that we can process RSS based on inner 5 tuples or
3 tuples of all the cases below of GTP packets:
        1. ipv4 (outer) + ipv4 (inner)
        2. ipv4 (outer) + ipv6 (inner)
        3. ipv6 (outer) + ipv4 (inner)
        4. ipv6 (outer) + ipv6 (inner)

Signed-off-by: Junfeng Guo <junfeng.guo@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>

net/ice: fix protocol header for PPPoE

When adding a RSS rule with pattern MAC_PPPOE_IPV4_UDP and input
set SRC/DST IPV4, because of incomplete protocol header fields,
the rule can't do hash with inner src/dst ipv4. PPPOE_IPV4_TCP/SCTP
and PPPOE_IPV6_UDP/TCP/SCTP also have similar issues. This patch
complements protocol header fields for PPPOE data packets.

Fixes: 0b952714e9c1 ("net/ice: refactor PF hash flow")
Signed-off-by: Simei Su <simei.su@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>

net/ixgbe/base: update version

Update base code version in readme.

Signed-off-by: Guinan Sun <guinanx.sun@intel.com>
Reviewed-by: Wei Zhao <wei.zhao1@intel.com>

net/ixgbe/base: check host interface return status

Writing to read-only fields returns a non-OK Return Status
for shadow RAM write command for X550.
This information was previously discarded.

Signed-off-by: Stanislaw Grzeszczak <stanislaw.a.grzeszczak@intel.com>
Signed-off-by: Guinan Sun <guinanx.sun@intel.com>
Reviewed-by: Wei Zhao <wei.zhao1@intel.com>

net/ixgbe/base: cleanup pre-processor tags

The codes has been exposed correctly, so remove pre-processor tags.

Signed-off-by: Guinan Sun <guinanx.sun@intel.com>
Reviewed-by: Wei Zhao <wei.zhao1@intel.com>

net/ixgbe/base: remove default advertising for x550 2.5G/5G

We are seeing interoperability issues with switches when 2.5G and 5G
in x550 are advertised by default, so default to off.

Signed-off-by: Todd Fujinaka <todd.fujinaka@intel.com>
Signed-off-by: Guinan Sun <guinanx.sun@intel.com>
Reviewed-by: Wei Zhao <wei.zhao1@intel.com>

net/ixgbe/base: add IPv6 mask for flow director

Write FDIRIP6M register to allow flow director filter
to set ipv6 rules without setting ipv6 source/destination address.

Signed-off-by: Piotr Skajewski <piotrx.skajewski@intel.com>
Signed-off-by: Guinan Sun <guinanx.sun@intel.com>
Reviewed-by: Wei Zhao <wei.zhao1@intel.com>

net/ixgbe/base: improve log about autoneg being disabled

On ESXi OS, when user disables auto negotiation, the following log
appears: "(unsupported) Flow control autoneg is disabled".
It is true that auto negotiation is disabled but it is
not necessarily true that it is not supported.

Signed-off-by: Jakub Chylkowski <jakubx.chylkowski@intel.com>
Signed-off-by: Guinan Sun <guinanx.sun@intel.com>
Reviewed-by: Wei Zhao <wei.zhao1@intel.com>

net/ixgbe/base: initialize data field in struct buffer

While sending request using ixgbe_hic_unlocked() the data field in
buffer struct is not used. It is set when the struct is overwritten by
FW to deliver the response. To not pass random data to FW the whole
structure should be zeroed before use.

Signed-off-by: Krzysztof Galazka <krzysztof.galazka@intel.com>
Signed-off-by: Piotr Pietruszewski <piotr.pietruszewski@intel.com>
Signed-off-by: Guinan Sun <guinanx.sun@intel.com>
Reviewed-by: Wei Zhao <wei.zhao1@intel.com>

net/ixgbe/base: remove log message FC autoneg

The function ixgbe_device_supports_autoneg_fc is checking whether
a particular device and medium configuration is supporting
Flow Control Autonegotiation. In case of non-support, the message
is always logged which is confusing.
The fix is removing unnecessary log entry.

Signed-off-by: Zalfresso-Jundzillo <marekx.zalfresso-jundzillo@intel.com>
Signed-off-by: Guinan Sun <guinanx.sun@intel.com>
Reviewed-by: Wei Zhao <wei.zhao1@intel.com>

net/ixgbe/base: remove unnecessary return value check

Remove unnecessary return value check.

Signed-off-by: Jakub Chylkowski <jakubx.chylkowski@intel.com>
Signed-off-by: Guinan Sun <guinanx.sun@intel.com>
Reviewed-by: Wei Zhao <wei.zhao1@intel.com>

net/ixgbe/base: add typecast for type mismatch

Add typecast for type mismatch.

Signed-off-by: Jakub Chylkowski <jakubx.chylkowski@intel.com>
Signed-off-by: Guinan Sun <guinanx.sun@intel.com>
Reviewed-by: Wei Zhao <wei.zhao1@intel.com>

net/ixgbe/base: create function to restart autoneg

This patch is for restarting auto negotiation on PHY.

Signed-off-by: Jakub Chylkowski <jakubx.chylkowski@intel.com>
Signed-off-by: Guinan Sun <guinanx.sun@intel.com>
Reviewed-by: Wei Zhao <wei.zhao1@intel.com>

net/ixgbe/base: move increments after evaluations

The retry variable was being incremented before it was evaluated by the
subsequent conditional against the maximum retries to figure out which
message to print.  So we'll move the increment op to the end.

Signed-off-by: Jeb Cramer <jeb.j.cramer@intel.com>
Signed-off-by: Guinan Sun <guinanx.sun@intel.com>
Reviewed-by: Wei Zhao <wei.zhao1@intel.com>

net/ixgbe/base: remove whitespace in function comments

Remove unnecessary extra whitespace on all function comments, replacing
' *  ' with ' * '.

This was done automatically via sed using the following transformation:

  sed 's/^ \*  / * /'

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Guinan Sun <guinanx.sun@intel.com>
Reviewed-by: Wei Zhao <wei.zhao1@intel.com>

net/ixgbe/base: cleanup spelling mistakes in comments

Several functions in the driver code have a weird function comment
formatting which uses two spaces instead of only one space for the main
function body.

This formatting will be mechanically fixed by sed in a future patch, but
doing so leads to some spelling warnings on that patch. Cleanup the
spelling mistakes that will be detected first. This way, it is easier to
verify the mechanical transformation done by sed in the following patch.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Guinan Sun <guinanx.sun@intel.com>
Reviewed-by: Wei Zhao <wei.zhao1@intel.com>

net/ixgbe/base: add register definitions for NVM update

Added additional register for X550 and above device family.

Signed-off-by: Piotr Skajewski <piotrx.skajewski@intel.com>
Signed-off-by: Guinan Sun <guinanx.sun@intel.com>
Reviewed-by: Wei Zhao <wei.zhao1@intel.com>

net/ixgbe/base: fix infinite recursion on PCIe link down

In some corner cases the functions ixgbe_clear_rar_generic and
ixgbe_clear_vmdq_generic may call one another leading to infinite
recursion.

When ixgbe_clear_vmdq_generic is called with IXGBE_CLEAR_VMDQ_ALL
flag, it's going to clear MPSAR registers, and proceed to call
ixgbe_clear_rar_generic, which in turn will clear the RAR registers,
and recursively call back ixgbe_clear_vmdq_generic. Normally, the
latter would detect that MPSAR registers have already been cleared
and terminate the recursion.

However, when PCIe link is down, and before the driver has had the
opportunity to shut itself down, all register reads return 0xFFFFFFFF,
and all register writes fail silently. In such case, because
ixgbe_clear_vmdq_generic blindly assumes that clearing MPSAR registers
succeeded, it's going to always call ixgbe_clear_rar_generic, which
in turn will always call back ixgbe_clear_vmdq_generic, creating
infinite recursion.

This patch re-reads MPSAR register values after they had been cleared.
In case of PCIe link failure, the values read will be non-zero, which
will terminate the recursion. On the other hand, under normal
circumstances the value read from MPSAR registers is going to be equal
to the value previously written, so this patch is expected not to cause
any regressions.

Fixes: af75078fece3 ("first public release")
Cc: stable@dpdk.org
Signed-off-by: Robert Konklewski <robertx.konklewski@intel.com>
Signed-off-by: Guinan Sun <guinanx.sun@intel.com>
Reviewed-by: Wei Zhao <wei.zhao1@intel.com>

net/ixgbe/base: fix x550em 10G NIC link status

With the NVM image for x550em XFI will not report
the auto-negotiation feature correctly. The auto-negotiation
should be "No" for supports and advertised items.
At the same time update speed makes it support 1G and 10G.

Fixes: 833df43399e7 ("net/ixgbe/base: add SGMII link for X550")
Cc: stable@dpdk.org
Signed-off-by: Piotr Skajewski <piotrx.skajewski@intel.com>
Signed-off-by: Guinan Sun <guinanx.sun@intel.com>
Reviewed-by: Wei Zhao <wei.zhao1@intel.com>

net/ixgbe/base: fix response to apply-update command

For the "Apply Update" command the firmware does not
given an response. For this command, success should
be return.

Signed-off-by: Mateusz Kowalski <mateusz.kowalski@intel.com>
Signed-off-by: Guinan Sun <guinanx.sun@intel.com>
Reviewed-by: Wei Zhao <wei.zhao1@intel.com>

net/ixgbe/base: fix host interface shadow RAM read

Host interface Shadow RAM Read (0x31) command response
buffer length should be stored in two bytes, instead of one byte.
This patch fixes it.

Fixes: e6102361b1d4 ("net/ixgbe/base: use 2 bytes for flash read command")
Cc: stable@dpdk.org
Signed-off-by: Mateusz Kowalski <mateusz.kowalski@intel.com>
Signed-off-by: Guinan Sun <guinanx.sun@intel.com>
Reviewed-by: Wei Zhao <wei.zhao1@intel.com>

net/qede: support registers dump

Add support for .get_reg eth_dev ops which will be used to collect the
firmware debug data.

PMD on detecting on some HW errors will collect the FW/HW Dump to a
buffer and then it will save it to a file implemented in
qede_save_fw_dump().

Dump file location and name:
Location: <RTE_SDK> or DPDK root
Name: qede_pmd_dump_mm-dd-yy_hh-mm-ss.bin

DPDK applications can initiate a debug data collection by invoking DPDK
library’s rte_eth_dev_get_reg_info() API. This API invokes .get_reg()
interface in the PMD.

PMD implementation of .get_reg() collects the FW/HW Dump, saves it to
data field of rte_dev_reg_info and passes it to the application. It’s
the responsibility of the application to save the FW/HW Dump to a file.
We recommendation using the file name format used by qede_save_fw_dump().

Signed-off-by: Rasesh Mody <rmody@marvell.com>
Signed-off-by: Igor Russkikh <irusskikh@marvell.com>

net/qede: add infrastructure for debug data collection

The patch adds QEDE PMD ops and APIs to calculate the size and collect
the debug dump for various firmware components. The patch adds new files
qede_debug.[ch] that has all the firmware debug data collection
infrastructure changes.

Signed-off-by: Rasesh Mody <rmody@marvell.com>
Signed-off-by: Igor Russkikh <irusskikh@marvell.com>

net/qede/base: add changes for debug data collection

This patch adds base driver APIs required for debug data collection.
It adds support for dumping internal lookup tables(ilt), reading nvram
image, register definitions.

Signed-off-by: Rasesh Mody <rmody@marvell.com>
Signed-off-by: Igor Russkikh <irusskikh@marvell.com>

net/qede/base: re-arrange few structures for DDC

This patch rearranges some of the base driver structures which will be
also used by debug data collection (DDC) implementation. It adds a new
file ecore_hsi_func_common.h with Physical, Virtual memory descriptors.

Signed-off-by: Rasesh Mody <rmody@marvell.com>
Signed-off-by: Igor Russkikh <irusskikh@marvell.com>

net/netvsc: detach external buffer on failure

When external buffer is used, driver should detach it if it doesn't make
it successfully to the queue.

Signed-off-by: Long Li <longli@microsoft.com>
Acked-by: Stephen Hemminger <stephen@networkplumber.org>

net/netvsc: fix underflow when Rx external mbuf

When rte_pktmbuf_attach_extbuf() is used, the driver should not decrease
the reference count in its callback function hn_rx_buf_free_cb, because
the reference count is already decreased by rte_pktmbuf. Doing it twice
may result in underflow and driver may never send an ack packet over
vmbus to host.

Also declares rxbuf_outstanding as atomic, because this value is shared
among all receive queues.

Fixes: 4e9c73e96e83 ("net/netvsc: add Hyper-V network device")
Cc: stable@dpdk.org
Signed-off-by: Long Li <longli@microsoft.com>
Acked-by: Stephen Hemminger <stephen@networkplumber.org>

net/bonding: fix socket ID check

Caught by code review, rte_eth_dev_socket_id() returns -1 on error.
The code should behave the same, but still, do not use LCORE_ID_ANY for
something that is not a lcore id.

Fixes: c15c5897340d ("net/bonding: avoid allocating mempool on unknown socket")
Cc: stable@dpdk.org
Signed-off-by: David Marchand <david.marchand@redhat.com>
Acked-by: Chas Williams <chas3@att.com>

net/bnxt: avoid hard coded values when reading counters

Instead of using hardcoded values for the byte/pkt value shifts/masks
to read from the HW counters, use the shift/mask values from the device
template params

Reviewed-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/bnxt: enable filter ctrl ops for port representor

Inorder to offload flows on the vfrep device, it must be
populated with rte_flow_ops.

This patch enables the same.

Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/bnxt: use VF VNIC if port action is for VF rep

Fix to use the vf's vnic port for ingress flows whose
port action is a vf rep port.

Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com>
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/bnxt: support src and dst port actions

Added support for set transport port source and destination
rewrite action items. This allows changing the tcp or udp
source/destination ports for a given flow.

Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com>
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
Reviewed-by: Mike Baucom <michael.baucom@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/bnxt: support NAT action items

Added support for set ipv4 address action items. It allows the source
or destination ip address to be changed for a given flow.

Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com>
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
Reviewed-by: Mike Baucom <michael.baucom@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/bnxt: add conditional opcodes for mapper result table

Added support for conditional mapper result opcodes. The conditional
opcodes allows to set the action details in hardware based on the
actions configured for the flow. This allows aggregation of multiple
templates.

Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com>
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
Reviewed-by: Mike Baucom <michael.baucom@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/bnxt: change port config for full offload

Added port configuration changes to support full offload
rules when VF representor ports are used. The direction of
the flow is determined using the configured direction and the
configured match and action ports of the flow create.

Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com>
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
Reviewed-by: Mike Baucom <michael.baucom@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/bnxt: enable PF and VF port action items

Added support for the PF and VF port action items in the flow
create. During flow create the output port action can now be specified
as PF or VF port and those ports are parsed accordingly and converted
to vnic or vport as per the flow direction.

Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com>
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
Reviewed-by: Mike Baucom <michael.baucom@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/bnxt: add VLAN tag count in computed field list

Added number of vlan tags in the computed field list so conditional
table execution could be done based on number of vlan tags in the
flow create.

Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com>
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: Mike Baucom <michael.baucom@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/bnxt: move VXLAN outer IP protocol ID in encap

The outer ip protocol was not encapsulated in the right location
when ip header is sent by the application. The order of encapsulation
has to be reversed.

Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com>
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
Reviewed-by: Mike Baucom <michael.baucom@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/bnxt: remove VNIC and vport bits from template match

Removed the vnic and vport bitmaps from template matching. It
is assumed that these will be populated implicitly and based
on the direction the appropriate action property shall be used.

Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com>
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
Reviewed-by: Mike Baucom <michael.baucom@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/bnxt: support VLAN push and pop actions

Add support for the vlan push and vlan pop actions

Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com>
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
Reviewed-by: Mike Baucom <michael.baucom@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/bnxt: cleanup and refactor session management

The return value of some functions is explicitly ignored
in cases where scope id may not be valid for internal EM
entries.
Additional minor refactoring and cleanups
- Change log level for some log messages to DEBUG instead of ERR.
- Check data size conformity and log appropriate message.

Signed-off-by: Michael Wildt <michael.wildt@broadcom.com>
Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com>
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: Randy Schacher <stuart.schacher@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/bnxt: add HWRM for global config

Change global cfg from tunneled to non-tunneled
HWRM cmds.

Signed-off-by: Jay Ding <jay.ding@broadcom.com>
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
Reviewed-by: Peter Spreadborough <peter.spreadborough@broadcom.com>
Reviewed-by: Randy Schacher <stuart.schacher@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/bnxt: update HWRM structures

Brought in the latest hsi_struct_def_dpdk.h in order to get
the TF global cfg set/get HWRM cmds.
HWRM API is now updated to version 1.10.1.48

Signed-off-by: Jay Ding <jay.ding@broadcom.com>
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
Reviewed-by: Randy Schacher <stuart.schacher@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/bnxt: support index opcode constant

Add support for index opcode constant so that
parif configuration could be constant value.

Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com>
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
Reviewed-by: Mike Baucom <michael.baucom@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/bnxt: validate port ID and PHY port parsing

Added validations to the ulp parser to validate the dpdk port id
and phy port index during the flow creation.

Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com>
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
Reviewed-by: Mike Baucom <michael.baucom@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/bnxt: support negative conditional opcodes

Added support for negative conditional opcodes in the
mapper processing.

Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com>
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
Reviewed-by: Mike Baucom <michael.baucom@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/bnxt: support VLAN header bitmap

Add support for the vlan headers in the matching of the flow
patterns.

Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com>
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
Reviewed-by: Mike Baucom <michael.baucom@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/bnxt: use source property for VXLAN encap/decap

The vxlan encap and decap flows need to allocate the source
record property and populate the action fields during the
flow parsing.

Signed-off-by: Kishore Padmanabha <kishore.padmanabha@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>

net/bnxt: support NEON

Add bnxt vector PMD support using NEON SIMD instructions.
Also update the 20.08 release notes with this information.

Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Reviewed-by: Ruifeng Wang <ruifeng.wang@arm.com>

net/bnxt: support Rx/Tx burst mode info

Retrieve burst mode options according to the selected Rx/Tx burst
function name.
Update 20.08 release notes with this information.

Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/bnxt: disable vector Rx for mark action

The bnxt vector mode receive handler does not support the rte_flow
'mark' action. Since we cannot know in advance whether this action
will be required, add support for dynamically switching from vector
to non-vector receive when the first flow create request with a
mark action is processed.

Fixes: 94eb699bc82e ("net/bnxt: support flow mark action")
Cc: stable@dpdk.org
Suggested-by: Thomas Monjalon <thomas@monjalon.net>
Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/dpaa2: support raw flow classification

Add support for raw flow, which can be used for any
protocol rules.

Signed-off-by: Nipun Gupta <nipun.gupta@nxp.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>

net/dpaa2: configure per-class distribution size

TC distribution size is set with dist_queues or

nb_rx_queues % dist_queues in order of TC priority.

Signed-off-by: Jun Yang <jun.yang@nxp.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>

net/dpaa2: support flow API FS miss action configuration

1) dpni_set_rx_hash_dist and dpni_set_rx_fs_dist used for TC
   configuration instead of dpni_set_rx_tc_dist. Otherwise,
   re-configuration of default TC of QoS fails.

2) Default miss action is to drop.  "export
   DPAA2_FLOW_CONTROL_MISS_FLOW=flow_id" is used receive the missed
   packets from flow with flow ID specified.

Signed-off-by: Jun Yang <jun.yang@nxp.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>

net/dpaa2: modify flow API QoS setup to follow FS setup

In HW/MC logical, QoS setup should follow FS setup.
In addition, Skip QoS setup if MAX TC number of DPNI is set 1.

Signed-off-by: Jun Yang <jun.yang@nxp.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>

net/dpaa2: add flow data sanity check

Check flow attributions and actions before creating flow.
Otherwise, the QoS table and FS table need to re-build
if checking fails.

Signed-off-by: Jun Yang <jun.yang@nxp.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>

net/dpaa2: support index of queue action for flow

Make more sense to use RXQ index for queue distribution

instead of flow ID.

Signed-off-by: Jun Yang <jun.yang@nxp.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>

net/dpaa2: support distribution size set on multiple TCs

Default distribution size of TC is 1, which is limited by MC. We have to
set the distribution size for each TC to support multiple RXQs per TC.

Signed-off-by: Jun Yang <jun.yang@nxp.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>

net/dpaa2: differentiate IPv4/IPv6 in IP flow setup

Discriminate between IPv4 and IPv6 in generic IP flow setup.

Signed-off-by: Jun Yang <jun.yang@nxp.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>

net/dpaa2: add logging of flow extracts and rules

This patch add support for logging the flow rules.

Signed-off-by: Jun Yang <jun.yang@nxp.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>

net/dpaa2: define size of table entry

If entry size is not bigger than 27, MC alloc one TCAM entry,
otherwise, alloc 2 TCAM entries.
Extracts size by HW must be not bigger than TCAM entry size(27 or 54).
So define the flow entry size as 54.

Signed-off-by: Jun Yang <jun.yang@nxp.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>

net/dpaa2: support QoS or FS table entry indexing

Calculate QoS/FS entry index by group and priority of flow.

1)The less index of entry, the higher priority of flow.
2)Verify if the flow with same group and priority has been added before
  creating flow.

Signed-off-by: Jun Yang <jun.yang@nxp.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>

net/dpaa2: free flow rule memory

Free rule memory when the flow is destroyed.

Signed-off-by: Jun Yang <jun.yang@nxp.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>

net/dpaa2: add sanity check for flow extracts

Define extracts support for each protocol and check the fields of each
pattern before building extracts of QoS/FS table.

Signed-off-by: Jun Yang <jun.yang@nxp.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>

net/dpaa2: support key extracts of flow API

1) Support QoS extracts and TC extracts for multiple TCs.

2) Protocol type of L2 extract is used to parse L3.
   Next protocol of L3 extract is used to parse L4.

3) generic IP key extracts instead of IPv4 and IPv6 respectively.

4) Special for IP address extracts:
   Put IP(v4/v6) address extract(s)/rule(s) at the end of extracts array
   to make rest fields at fixed position.

Signed-off-by: Jun Yang <jun.yang@nxp.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>

net/dpaa2: support dynamic flow control

Dynamic flow used instead of layout defined.

The actual key/mask size depends on protocols and(or) fields
of patterns specified.
Also, the key and mask should start from the beginning of IOVA.

Signed-off-by: Jun Yang <jun.yang@nxp.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>

bus/dpaa: enable set link status

Enabled set link status API to start/stop phy
device from application.

Signed-off-by: Rohit Raj <rohit.raj@nxp.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>