net/ice/base: fix packet type bitmap

Before this patch, IPv4 and UDP inner hash rule will be over
written by later rules after RSS initialization phase. This is
because the PTYPE bitmap table cover some PTYPEs belong to another
PTGs. And some PTYPEs are reserved. Remove these PTYPEs in TCP,
UDP, SCTP and ipv4 bitmap table.

Fixes: aa1cd410fa64 ("net/ice/base: add flow module")

Signed-off-by: Zhirun Yan <zhirun.yan@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

net/ice/base: fix alignment

As title says, fix an alignment issue.

Fixes: 51d04e4933e3 ("net/ice/base: add flexible pipeline module")
Cc: stable@dpdk.org
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

net/ice/base: use bitmap copy where appropriate

ice_cp_bitmap() already exists and should be used instead of
using ice_memcpy().  Note, there are a couple comments that suggest
using a bitmap-specific copy function, but those are not correct
since the source block of memory is not a bitmap.

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

net/ice/base: remove unnecessary error log

Remove the error log message when attempting to download a
package that has an unsupported version.

Signed-off-by: Dan Nowlin <dan.nowlin@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

net/ice/base: fix 4 bytes alignment for PPPoE dummy packet

Add two bytes to meet the requirement of 4 bytes alignment for dummy
packet for creating switch rule for PPPoE.

Fixes: d1c2f76b440a ("net/ice/base: support GTP and PPPoE protocols")

Signed-off-by: Junfeng Guo <junfeng.guo@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

net/ice/base: search field vector indices for result slots

Previously, switch code would use only pre-reserved index
slots at the end of each field vector for recipe result index
locations. This patch adds code that detects other internal
empty index slots that could potentially be used. For each
recipe that is added, a determ ination is made as to whether
any of these additional index slots alige with all the profiles
selected for the recipe; if alignment is achieved, then these
result index slots can be used.

Signed-off-by: Dan Nowlin <dan.nowlin@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

net/ice/base: remove unused DDP package macros

Macros no longer be used and can be removed

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

net/ice/base: fix segment in remove existing RSS rule

Before this patch, RSS tunneled rules can not be destroyed at runtime.
This is because it can not find the existing matching profile for tunnels.
segs[0] should always be zero and all matched, segs[1] for inner part.
It only construct one segment.

This patch modifies construct segment in ice_rem_rss_cfg_sync() to
match ice_add_rss_cfg_sync().

Fixes: b7d34ccc47b5 ("net/ice/base: packet encapsulation for RSS")

Signed-off-by: Zhirun Yan <zhirun.yan@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

net/ice/base: fix bitmap for TCP in RSS

Before this patch, if set rule for IPv4 first and then set rule
for TCP with IPv4. The first rule for inner IP will be overwritten
by TCP rule. This is because MAC_IPV6_TUN_MAC_IPV4_PAY using the
same ptgs PTG_TUN_INNER_IPV4_OTHER with MAC_IPV4_TUN_MAC_IPV4_PAY,
this ptype should not in TCP bitmap.
Remove this bit in TCP bitmap.

Fixes: aa1cd410fa64 ("net/ice/base: add flow module")
Cc: stable@dpdk.org
Signed-off-by: Zhirun Yan <zhirun.yan@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

net/ice/base: support FDIR for GTPU QFI field

Add GTPU qfi field support for flow director. Note that for GTPU pkt,
only qfi field (6 bits) can be set for FD. The supported GTPU pkts are
defined as:
ICE_FLTR_PTYPE_NONF_IPV4_GTPU_IPV4_UDP,
ICE_FLTR_PTYPE_NONF_IPV4_GTPU_IPV4_TCP,
ICE_FLTR_PTYPE_NONF_IPV4_GTPU_IPV4_ICMP,
ICE_FLTR_PTYPE_NONF_IPV4_GTPU_IPV4_OTHER (FRAG and PAY belong to this)

Signed-off-by: Junfeng Guo <junfeng.guo@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

net/ice/base: replace alloc-followed-by-copy with memdup

ice_memdup() is preferred over an alloc immediately followed by a copy.

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

net/ice/base: enable symmetric hash for RSS

Add parameter "symm" to rss configuration APIs.
When symm is 1, Symmetric Teoplitz Hash can be enabled by
configuring GLQF_HSYMM properly.

NOTE:
Symmetric Teoplitz hash will work only if hash schema of
VSIQF_HASH_CTL be configured to 01b and it is assumed be enabled
in PMD.

Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

net/ice/base: update FW API minor version

Update FW API minor version to align to current value advertised
by FW in NVM images.

Signed-off-by: Kevin Scott <kevin.c.scott@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

net/ice/base: support destination MAC field for FDIR

Add dest MAC address support so that this field can be matched when
we set Flow Director filter with dst addr for MAC.

Signed-off-by: Junfeng Guo <junfeng.guo@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

net/ice/base: enable setting up FDIR counters

Enable getting value from input to set up flow director counters,
so that the FDIR counters can count none, packets only, bytes only
or both packets and bytes as demanded.

Signed-off-by: Junfeng Guo <junfeng.guo@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

net/ice/base: enable FDIR queue region

Add fdir queue region support.

Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

net/ice/base: enable RSS for PPPoE with SCTP

Add two ptypes(MAC_PPPOE_IPV4_SCTP and MAC_PPPOE_IPV6_SCTP) in
sctp ptype bitmap to enable rss.

Signed-off-by: Zhirun Yan <zhirun.yan@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

net/ice/base: associate switch recipe to profiles

Properly associate switch recipes to profiles. Previous code was
using the wrong bitfield for updating the associations, which was
causing other PFs to not properly identify and use existing
recipes. This sometimes resulted in rules not being added when it
should have been possible.

Signed-off-by: Dan Nowlin <dan.nowlin@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

net/ice/base: refactor data structure

When declaring the ice_prot_ext, and ice_prot_id_tbl structure,
we can use a fixed length array instead of a variable length one
which helps us catch future code changes that might desynchronize
the enum ice_protocol_type and the structs.  This change also
necessitates removing the last member of the structs which was
just there to be a placeholder.

Also reorder the ice_prot_ext struct to match the ordering in the
associated enum ice_protocol_type.

Signed-off-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

net/ice/base: support switch rule about VLAN PPPoE

Add support for switch rule about single-VLAN-PPPoE. Note that double
VLAN is not supported by the hardware at this point, therefore only
single-VLAN support for PPPoE is added.

Signed-off-by: Junfeng Guo <junfeng.guo@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

net/ice/base: remove RSS code as iavf host

The DPDK PF doesn't support SRIOV so remove the related iavf host
code.

Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

net/ice/base: remove null check of port info

The code in ice_sched_cleanup_all checks whether the port info is NULL
prior to calling ice_sched_clear_port.

More importantly, it also checks whether the port structure has been
initialized by checking its port_state field as well.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

net/ice/base: propagate errors from functions

There could be an error returned from ice_fill_adv_dummy_packet()
so we need to propagate that to the caller. Additionally, the
call to ice_flow_xtract_pkt_flags() could also return an error so
we need to propagate it as well.

Also add in the correct offsets for GENEVE and VXLAN_GPE to the
dummy packets.

Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

net/ice/base: fix mask for checking protocol header

Before this patch, the logic of protocol header checking only support
non-tunneled packet. This patch remove the inner protocol in L3/L4 RSS
seg hdr mask and change the protocol header validation to reflect this.
So, for ice_add_rss_cfg(), the last parameter addl_hdrs could specify
the protocol header for tunnel.

Signed-off-by: Zhirun Yan <zhirun.yan@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

net/ice/base: add non-word aligned IPv6 field support

Add non-word aligned field support for IPv6 with hlim, tc and proto.
All these fields are one byte within one word. In order to match
bytes within the IPv6 header for flow director we need to use a mask.

Signed-off-by: Junfeng Guo <junfeng.guo@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

net/ice/base: add non-word aligned IPv4 field support

Add non-word aligned field support for ipv4 with ttl, tos and proto.
All these fields are one byte within one word. In order to match
bytes within the IPv4 header for flow director we need to use a mask.

Signed-off-by: Junfeng Guo <junfeng.guo@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

net/ice/base: support tunnel packets

Add VXLAN tunnel training packets to flow director and change the
interface to support tunnel packets.

Signed-off-by: Henry Tieman <henry.w.tieman@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

net/ice/base: remove redundant empty lines

Remove redundant empty lines

Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Qiming Yang <qiming.yang@intel.com>

examples/ipv4_multicast: check allmulticast enable status

rte_eth_allmulticast_enable()/rte_eth_allmulticast_disable() return
value was changed from void to int, so this patch modify usage
of these functions across examples/ipv4_multicast
according to new return type.

Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>

app/testpmd: check code of allmulticast mode switch

rte_eth_allmulticast_enable()/rte_eth_allmulticast_disable() return
value was changed from void to int, so this patch modify usage
of these functions across app/test-pmd
according to new return type.

Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>

ethdev: do nothing if all-multicast mode is applied again

Since driver callbacks return status code now, there is no necessity
to enable or disable all-multicast mode once again if it is already
successfully enabled or disabled.

Configuration restore at startup tries to ensure that configured
all-multicast mode is applied and start will return error if it fails.

Also it avoids theoretical cases when already configured all-multicast
mode is applied once again and fails. In this cases it is unclear
which value should be reported on get (configured or opposite).

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>

ethdev: change allmulticast callbacks to return status

Enabling/disabling of allmulticast mode is not always successful and
it should be taken into account to be able to handle it properly.

When correct return status is unclear from driver code, -EAGAIN is used.

Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Acked-by: Hyong Youb Kim <hyonkim@cisco.com>

net/bonding: check code of allmulticast mode switch

rte_eth_allmulticast_enable()/rte_eth_allmulticast_disable() return
value was changed from void to int, so this patch modify usage
of these functions across net/bonding
according to new return type.

Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>

net/failsafe: check code of allmulticast mode switch

rte_eth_allmulticast_enable()/rte_eth_allmulticast_disable() return
value was changed from void to int, so this patch modify usage
of these functions across net/failsafe according to new return type.

Try to keep all-multicast mode consistent across all active
devices in the case of failure.

Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Acked-by: Gaetan Rivet <gaetan.rivet@6wind.com>

ethdev: change allmulticast mode API to return errors

Change rte_eth_allmulticast_enable()/rte_eth_allmulticast_disable()
return value from void to int and return negative errno values
in case of error conditions.
Modify usage of these functions across the ethdev according
to new return type.

Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>

ethdev: change owner delete function to return int

Change rte_eth_dev_owner_delete() return value from void to int
and return negative errno values in case of error conditions.

Right now there is only one error case for rte_eth_dev_owner_delete() -
invalid owner, but it still makes sense to return error to catch bugs
in the code which uses the function.

Also update the usage of the function in drivers/netvsc
according to the new return type.

Signed-off-by: Igor Romanov <igor.romanov@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>

examples: check status of getting MAC address

The return value of rte_eth_macaddr_get() was changed from void to int.
Update the usage of the functions according to the new return type.

Signed-off-by: Igor Romanov <igor.romanov@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>

app: check status of getting MAC address

The return value of rte_eth_macaddr_get() was changed from void to int.
Update the usage of the functions according to the new return type.

Signed-off-by: Igor Romanov <igor.romanov@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>

app/testpmd: check status of getting MAC address

Add a wrapper for rte_eth_macaddr_get() that prints an
error and returns a status code if the function fails.

Signed-off-by: Igor Romanov <igor.romanov@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>

ethdev: change MAC address get function to return int

Change rte_eth_macaddr_get() return value from void to int
and return negative errno values in case of error conditions.

Signed-off-by: Igor Romanov <igor.romanov@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>

examples: check status of getting link info

The return value of rte_eth_link_get() and rte_eth_link_get_nowait()
was changed from void to int. Update the usage of the functions
according to the new return type.

Signed-off-by: Igor Romanov <igor.romanov@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>

app: check status of getting link info

The return value of rte_eth_link_get() and rte_eth_link_get_nowait()
was changed from void to int. Update the usage of the functions
according to the new return type.

Signed-off-by: Igor Romanov <igor.romanov@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>

app/pipeline: check status of getting link info

The return value of rte_eth_link_get_nowait() was changed from void to int.
Update the usage of the function according to the new return type.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>

net/memif: check status of getting link info

The return value of rte_eth_link_get() was changed from void to int.
Update the usage of the function according to the new return type.

Just log error if something goes wrong.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>

net/ixgbe: check status of getting link info

The return value of rte_eth_link_get() and rte_eth_link_get_nowait()
was changed from void to int. Update the usage of the functions
according to the new return type.

Signed-off-by: Igor Romanov <igor.romanov@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>

net/bonding: check status of getting link info

The return value of rte_eth_link_get() and rte_eth_link_get_nowait()
was changed from void to int. Update the usage of the functions
according to the new return type.

Signed-off-by: Igor Romanov <igor.romanov@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>

app/testpmd: check status of getting link info

Add a wrapper for rte_eth_eth_link_get_nowait() that prints an
error and returns a status code if the function fails.

Signed-off-by: Igor Romanov <igor.romanov@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>

ethdev: change link status get functions to return int

Change rte_eth_link_get() and rte_eth_link_get_nowait() return value
from void to int and return negative errno values in case of error
conditions.

Return value of link_update callback is ignored since the callback
returns not errors but whether link up status has changed or not.

Signed-off-by: Igor Romanov <igor.romanov@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>

net/bonding: fix link speed update in broadcast mode

Fix the issue that the link speed of the bond device was set to the
link speed of the first active slave in broadcast mode.

Set the link speed of the bond device to the minimum value across
all of the slaves in that case.

Fixes: deba8a2f8b0b ("net/bonding: fix link properties management")
Cc: stable@dpdk.org
Signed-off-by: Igor Romanov <igor.romanov@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Acked-by: Chas Williams <chas3@att.com>

ethdev: make stats and xstats reset callbacks return int

Change return value of the callbacks from void to int. Make
implementations across all drivers return negative errno
values in case of error conditions.

Both callbacks are updated together because a large number of
drivers assign the same function to both callbacks.

Signed-off-by: Igor Romanov <igor.romanov@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

ethdev: change xstats reset function to return int

Change rte_eth_xstats_reset() return value from void to int and
return negative errno values in case of error conditions.

Signed-off-by: Igor Romanov <igor.romanov@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

examples: take promiscuous mode switch result into account

rte_eth_promiscuous_enable()/rte_eth_promiscuous_disable() return
value was changed from void to int, so this patch modify usage
of these functions across examples according to new return type.

Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>

app: check code of promiscuous mode switch

rte_eth_promiscuous_enable()/rte_eth_promiscuous_disable() return
value was changed from void to int, so this patch modify usage
of these functions across apps according to new return type.

Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>

app/testpmd: check code of promiscuous mode switch

rte_eth_promiscuous_enable()/rte_eth_promiscuous_disable() return
value was changed from void to int, so this patch modify usage
of these functions across app/testpmd
according to new return type.

Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Acked-by: Bernard Iremonger <bernard.iremonger@intel.com>

kni: check code of promiscuous mode switch

rte_eth_promiscuous_enable()/rte_eth_promiscuous_disable() return
value was changed from void to int, so modify usage of these
functions across lib/librte_kni according to new return type.

Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>

ethdev: do nothing if promiscuous mode is applied again

Since driver callbacks return status code now, there is no necessity
to enable or disable promiscuous mode once again if it is already
successfully enabled or disabled.

Configuration restore at startup tries to ensure that configured
promiscuous mode is applied and start will return error if it fails.

Also it avoids theoretical cases when already configured promiscuous
mode is applied once again and fails. In this cases it is unclear
which value should be reported on get (configured or opposite).

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>

ethdev: change promiscuous callbacks to return status

Enabling/disabling of promiscuous mode is not always successful and
it should be taken into account to be able to handle it properly.

When correct return status is unclear from driver code, -EAGAIN is used.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Acked-by: Matan Azrad <matan@mellanox.com>
Acked-by: Hyong Youb Kim <hyonkim@cisco.com>

net/bonding: check code of promiscuous mode switch

rte_eth_promiscuous_enable()/rte_eth_promiscuous_disable() return
value was changed from void to int, so this patch modify usage
of these functions across net/bonding according to new return type.

Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>

net/failsafe: check code of promiscuous mode switch

rte_eth_promiscuous_enable()/rte_eth_promiscuous_disable() return
value was changed from void to int, so this patch modify usage
of these functions across net/failsafe according to new return type.

Try to keep promiscuous mode consistent across all active
devices in the case of failure.

Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Acked-by: Gaetan Rivet <gaetan.rivet@6wind.com>

ethdev: change promiscuous mode controllers to return errors

Change rte_eth_promiscuous_enable()/rte_eth_promiscuous_disable()
return value from void to int and return negative errno values
in case of error conditions.
Modify usage of these functions across the ethdev according
to new return type.

Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>

doc: cleanup license in bnx2x guide

Removed redundant BSD boilerplate text from bnx2x guide.

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

net/octeontx2: add packet type translation for ICMP6

Update ptype translation logic with ICMP6
in both outer and inner layers.

Signed-off-by: Nithin Dabilpuram <ndabilpuram@marvell.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>

net/octeontx2: fix packet type translation

Extract and use layer type LB..LE for non-tunnel ptype and
LF..LH as tunnel ptype translation.

Fixes: 6e892eabce11 ("net/octeontx2: support packet type")
Cc: stable@dpdk.org
Signed-off-by: Nithin Dabilpuram <ndabilpuram@marvell.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>

net/octeontx2: fix port Tx VLAN insertion

This patch fixes configuration for port Tx vlan insertion feature.
As per latest Tx KEX profile, pf_func should be at
KW[0](47:32) and not KW[1](63:48).

Fixes: 168c59cfe42b ("net/octeontx2: add flow MCAM utility functions")
Cc: stable@dpdk.org
Signed-off-by: Nithin Dabilpuram <ndabilpuram@marvell.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>

net/octeontx2: fix packet type get overflow

The function `rte_eth_dev_get_supported_ptypes` expects the underlying
ethernet device to return array of supported ptypes. The ethernet device
needs to set `RTE_PTYPE_UNKNOWN` as the last element which signifies
thats its the end of the ptype array.
Else the function `rte_eth_dev_get_supported_ptypes` might overflow.

Fixes: 6e892eabce11 ("net/octeontx2: support packet type")
Cc: stable@dpdk.org
Signed-off-by: Pavan Nikhilesh <pbhagavatula@marvell.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>

doc: update octeontx2 supported SoCs

CNF95XX SoC part is not available for general availability.
Update the documentation to reflect the status.

Cc: stable@dpdk.org
Signed-off-by: Jerin Jacob <jerinj@marvell.com>

vhost: fix slave request fd leak

We need to close the old slave request fd if any first
before taking the new one.

Fixes: 275c3f944730 ("vhost: support slave requests channel")
Cc: stable@dpdk.org
Signed-off-by: Tiwei Bie <tiwei.bie@intel.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>

vhost: add device op when notification to guest is sent

This patch adds an operation callback which gets called every time
the library is waking up the guest trough an eventfd_write() call.

This can be used by 3rd party application, like OVS, to track the
number of times interrupts where generated. This might be of
interest to find out system-call were called in the fast path.

Signed-off-by: Eelco Chaudron <echaudro@redhat.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>

net/virtio: add Altivec Rx

Added the file virtio_rxtx_simple_altivec.c which implements Altivec
code for the virtio vectorized RX functions. Updated the various build
files.

Signed-off-by: David Christensen <drc@linux.vnet.ibm.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>

net/ice: support device-specific DDP package loading

This patch adds the feature that supports loading DDP package
according to the device serial number. Prior to loading the
default DDP package (ice.pkg), the driver will check for the
presence of a device-specific DDP package with the name containing
64-bit PCIe Device Serial Number (ice-xxxxxxxxxxxxxxxx.pkg)
during initialization. Users can use "lspci -vs" to get the device
serial number.
The pkg search path are /lib/firmware/updates/intel/ice/ddp/
and /lib/firmware/intel/ice/ddp/. If the package exists,
the driver will download it to the device instead of the default
one. The loaded package type (OS default and COMMS) will be
stored in ice_adapter->active_pkg_type. The package version is
stored in ice_hw->active_pkg_ver.

Signed-off-by: Ting Xu <ting.xu@intel.com>
Reviewed-by: Xiaolong Ye <xiaolong.ye@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>

net/ice/base: ignore inverse switch recipes

When looking for an existing recipes, never choose an inverse
recipe as these are used for anti-spoofing. Choosing inverse
recipes for source MAC address rules was causing errors while
adding the rule.

Signed-off-by: Dan Nowlin <dan.nowlin@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: update profile to recipe bitmap array

Correctly update profile to recipe bitmap array after adding and
associating recipes. This fixes an issue where determining unused
recipe result index slots was incorrect.

Signed-off-by: Dan Nowlin <dan.nowlin@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: add GENEVE offset

Add Geneve offset for tunneled packets to allow dummy packets to be
properly created.

Signed-off-by: Doug Dziggel <douglas.a.dziggel@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: enable RSS with ether layer for PPPoE

Add these two ptype(MAC_PPPOD_PAY and MAC_PPPOE_PAY) in outer mac
bitmap, so it can hash for outer mac.

Signed-off-by: Zhirun Yan <zhirun.yan@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: remove Rx flex descriptor programming

Removing Rx flex descriptor metadata and flag programming from the code,
these registers cannot be written to as they are read only.
The programming for all fields per RxDID is now handled differently.

Signed-off-by: Vignesh Sridhar <vignesh.sridhar@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: add AQC get link topology handle support

Add AQC get link topology handle support. This is needed to determine
Direct Attach (DA) or backplane media type for PHY types that support
either. Get link topology handle cage node type request can be used to
determine if a cage is present or not. If a cage is present for PHY
types that supports both DA and backplane media type, then the media
type is DA, else the media type is backplane.

Signed-off-by: Paul Greenwalt <paul.greenwalt@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: delay less

Shorten the delay for SQ responses, but increase the number of loops.
Max delay time is unchanged, but some operations complete much more
quickly.

In the process, add a new define to make the delay count and delay time
more explicit, and simplify the code so it's the same for both switch
and NIC mode. Add comments to make things more explicit.

Signed-off-by: Mitch Williams <mitch.a.williams@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: replace open-code duplication

Use BIT_ULL() instead of ICE_FLOW_HASH_FLD() which does the same bit shift
operation.

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: update to register definition file

Added register definitions for GL_MDCK_TX_TDPU and GL_MDET_TX_TDPU.

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: correct abbreviations

Correct abbreviation issues found by running abbrevcheck.

Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: remove unnecessary if branch

We are already in the branch "if (fm_list->vsi_count == 1)"
no need to exit and re-enter.

Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: remove unnecessary dummy packet finding

We don't need to find a dummy packet when removing a rule so remove
the call to get the dummy packet. This also obviates some variables
so remove them also.

Also reduce the scope of rule_buf_sz.

Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: update switch training packets with open ports

This patch updates UDP tunneled training packets with an appropriate
UDP dest port. For the correct profile to be chosen, an open tunnel
port must be included in the training packet.

Added GENEVE tunnel labels in the test package in order to test GENEVE
tunnel rule creation.

Signed-off-by: Dan Nowlin <dan.nowlin@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: maximize switch recipe words per line

Remove grouping rules to maximize the number of words placed into
a recipe line. This will allow more recipes to be added by reducing
the number of result indices required.

Signed-off-by: Dan Nowlin <dan.nowlin@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: refactor removal of VLAN promiscuous rules

Currently ice_clear_vsi_promisc() detects if the VLAN ID sent is not 0
and sets the recipe_id to ICE_SW_LKUP_PROMISC_VLAN in that case and
ICE_SW_LKUP_PROMISC if the VLAN_ID is 0. However this doesn't allow VLAN
0 promiscuous rules to be removed, but they can be added. Fix this by
checking if the promisc_mask contains ICE_PROMISC_VLAN_RX or
ICE_PROMISC_VLAN_TX. This change was made to match what is being done
for ice_set_vsi_promisc().

Signed-off-by: Brett Creeley <brett.creeley@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: fix flag settings in AQ call

Removed setting Read flag in the Get Allocated Resource Descriptors AQ
command (0x020A). The read flag is not required for this command and
causes the FW to return an error.

Fixes: d781ccbdd15d ("net/ice/base: add functions to get allocated resources")
Cc: stable@dpdk.org
Signed-off-by: Dan Nowlin <dan.nowlin@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: remove unnecessary conditional check

There is no reason to do this conditional check before the assignment so
simply remove it.

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: add RSS support for PPPoE and GTPU

Added RSS support for PPPoE with inner TCP and UDP.
Added RSS support for GTPU with IPv4 and IPv6.

Signed-off-by: Zhirun Yan <zhirun.yan@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: packet encapsulation for RSS

RSS configurations calls currently support only non-tunneled packets,
which uses outer-first IPs. Adding another call to configure RSS for
tunneled packets which will enable hashing on inner-last IP. RSS hash
will only be done on the innermost IP.
- Update RSS configuration calls to take packet segment count as input
  argument
- Update flow profile ID to include identifier for encapsulated packet
  [0:31] : Packet match fields
  [32:62]: Protocol header header
  [63]: Encapsulation flag, 0 if non-tunneled, 1 if tunneled

Signed-off-by: Vignesh Sridhar <vignesh.sridhar@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: fix for RSS hash on inner UDP port

Before this patch, only outer IP with inner UDP port can appear in
RSS hash, because the extraction sequence uses outer IP protocol
ID with Inner UDP protocol ID. ICE_PROT_UDP_OF always extracts the
TUNNELED UDP port values (i.e., 4789 for VXLAN). ICE_PROT_UDP_IL_OR_S
will extract NON-TUNNELED UDP port or inner UDP port.

Fixes: aa1cd410fa64 ("net/ice/base: add flow module")
Cc: stable@dpdk.org
Signed-off-by: Zhirun Yan <zhirun.yan@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: consolidate packet type group

This patch is an optimization to decrease the number of TCAM entries
used in the profile blocks, especially for RSS. To be most effective
this will also require a package change in order to decrease the number
of PTYPE groups necessary to program RSS, FD and ACL rules.

Change required package version to 1.3.

Signed-off-by: Dan Nowlin <dan.nowlin@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: add routine for tunnel port query

Add ice_get_open_tunnel_port routine, which can be used to find
an open tunnel port for creating switch and flow director training
packets.

Signed-off-by: Dan Nowlin <dan.nowlin@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: move and add some help function and macros

The patch move ice_ilog2 and ice_is_pow2 to ice_type.h, also add help
macro ROUND_UP and IS_ETHER_ADDR_EQUAL.

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: improve switch chained recipe

When updating switch database with chained recipes from
firmware, where a null pointer derefence was occurring due to looking
into a sub-recipe entry which is not filled in.

Signed-off-by: Dan Nowlin <dan.nowlin@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: update UDP tunnel switch training packets

Change UDP tunnel training packets for the switch to work for both
Geneve and VXLAN cases.

Signed-off-by: Dan Nowlin <dan.nowlin@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: return switch error on invalid match criteria

Modify ice_add_adv_rule to return an error when an invalid match
criteria is requested by the caller. This happens when the protocol
and offset pair is not supported by the package.

This change required a fix for the offset in the VXLAN GPE header,
and also found an issue with NVGRE where the package does not seem
to extract the correct offsets.

Signed-off-by: Dan Nowlin <dan.nowlin@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: associate recipes by profile type

Change recipe to profile association logic to determine the profile type
to determine if a profile is compatible with the rule being added.

Signed-off-by: Dan Nowlin <dan.nowlin@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: support not locking sideband queue

For certain PTP clock adjustments, there is a use case for locking the
sideband queue at a higher level and performing an atomic series of
operations while the sideband queue is locked.  To accommodate this use
case, split ice_sw_send_cmd() into a version that takes the lock and a
version that does not.

Signed-off-by: Ben Shelton <benjamin.h.shelton@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: add function to configure Tx AQ command

This function is needed by the driver to move PSM leaf nodes to a new
parent.

Additionally, add struct ice_aqc_move_txqs to struct ice_aq_desc so
driver code can access it.

Signed-off-by: Ben Shelton <benjamin.h.shelton@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>

net/ice/base: initialize driver NVM data earlier

The driver detects and warns about FW rollback mode before getting the NVM
version (through ice_init_nvm) which results in "0.00 0x0 0.0.0" being
printed. Fix this by calling ice_init_nvm earlier.

This could have been fixed by moving ice_print_rollback_msg as well, but
it made more sense to move ice_init_nvm closer to the flow that gets the
FW version information (i.e. control queue init flow).

Signed-off-by: Anirudh Venkataramanan <anirudh.venkataramanan@intel.com>
Signed-off-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com>
Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>