net/qede: fix port (re)configuration

Some applications set port configuration params like promisc mode
before calling dev_start(). This config results in a firmware exception
since this operation internally translates to sending of VPORT-UPDATE
before VPORT-START ramrod which is considered illegal from firmware
standpoint. So the fix is to send VPORT-START ramrod sooner
in dev_configure() rather than deferring it to dev_start().
This requires a bit of reshuffling in the code to move sending of
VPORT-START from qede_start_queues() to qede_dev_configure()
and VPORT-STOP from qede_stop_queues() to qede_dev_stop().

This sequence change also exposes a flaw in the port restart
flows where the fastpath resource allocation routine qede_init_fp()
functionalities need to be split, so that appropriate action is taken
based on the current port state. Eg: Do not re-initialize the status
block in a port restart case. This change ensures port start/stop
can be paired.

A new port state QEDE_DEV_CONFIG is added to distinguish between
port started from scratch vs port requiring a reconfig (like MTU).
The function qede_config_rx_mode() is removed since the individual
port config will be replayed anyways on a restart.

Fixes: 2ea6f76aff40 ("qede: add core driver")

Signed-off-by: Harish Patil <harish.patil@qlogic.com>

net/qede: support unequal number of Rx/Tx queues

Previous release of the qede PMD had a limitation that the
driver expects the number of tx and rx queues to be the same.
This patch fixes this issue by making appropriate changes in
control and data path.

Fixes: 2ea6f76aff40 ("qede: add core driver")

Signed-off-by: Sony Chacko <sony.chacko@qlogic.com>

net/qede/base: support MFW crash dump

Add support for management firmware(MFW) crash dump collection.

Signed-off-by: Rasesh Mody <rasesh.mody@qlogic.com>

net/qede/base: enhance comments

Comment additions and modifications

Signed-off-by: Rasesh Mody <rasesh.mody@qlogic.com>

net/qede/base: rename pmm prefix to eth

Renamed following to match with HSI changes
 - PMM_* to ETH_*
 - pmm_* to eth_*

Signed-off-by: Rasesh Mody <rasesh.mody@qlogic.com>

net/qede/base: update

This patch updates the base driver and incorporates necessary changes
required to bring in the new firmware 8.10.9.0.

In addition, it would allow driver to add new functionalities that might
be needed in future.

Signed-off-by: Rasesh Mody <rasesh.mody@qlogic.com>

net/qede/base: add selftest and query sensor info

This patch adds API support for NIC selftests (BIST) and APIs to retrieve
GPIO info, sensor data like temperature, MBA versions, ECC events etc.

Signed-off-by: Harish Patil <harish.patil@qlogic.com>

net/qede/base: serialize access to MFW mbox

Add ecore_mcp_mb_lock() and ecore_mcp_mb_unlock() APIs to ensure
a single thread is accessing MFW mailbox.

Signed-off-by: Harish Patil <harish.patil@qlogic.com>

net/qede/base: change signature of MCP command

Change ecore_mcp_cmd_and_union() to accept pointer to a structure rather
than accepting multiple arguments. A new struct ecore_mcp_mb_params is
added for that purpose. Also make this function static. This change is
mostly keeping in mind the future requests which needs additional
arguments.

Signed-off-by: Harish Patil <harish.patil@qlogic.com>

net/qede/base: fix 32-bit build

Fix 32 bit compilation for gcc version 4.3.4.

Fixes: ec94dbc57362 ("qede: add base driver")

Signed-off-by: Rasesh Mody <rasesh.mody@qlogic.com>

net/qede/base: update formatting and comments

Change details:
 - adds new comments
 - modifies some of the existing comments
 - abstract code into macros
 - split long lines

Signed-off-by: Rasesh Mody <rasesh.mody@qlogic.com>

net/qede/base: add attention formatting string

In case of attention from a signal that's represented by multiple bits
in misc AEU, add the format string which is populated with proper index
and resulting prints will show string as a prefix.

Signed-off-by: Rasesh Mody <rasesh.mody@qlogic.com>

net/qede/base: add HSI changes and register defines

- add the hardware software interface(HSI) changes
- add register definitions

These will be required for 8.10.9.0 FW upgrade.

Signed-off-by: Rasesh Mody <rasesh.mody@qlogic.com>

net/qede: use firmware config macros as needed

Replaced CONFIG_QED_BINARY_FW with CONFIG_ECORE_BINARY_FW.
Use CONFIG_ECORE_BINARY_FW and CONFIG_ECORE_ZIPPED_FW defines as
required.

Signed-off-by: Rasesh Mody <rasesh.mody@qlogic.com>

net/qede/base: formatting changes

Fixes white spaces and tabs.

Signed-off-by: Rasesh Mody <rasesh.mody@qlogic.com>

net/qede/base: add new files and rearrange the code

Added ecore_hsi_debug_tools.h, ecore_hsi_init_func.h,
ecore_hsi_init_tool.h files. Rearranged code from ecore_hsi_common.h and
ecore_hsi_tools.h to the new files. Removed unused code.

Signed-off-by: Rasesh Mody <rasesh.mody@qlogic.com>

net/mlx5: fix initialization in secondary process

The changes introduced by previous commits (ones in fixes lines) made
secondaries attempt to reinitialize the Tx queue structures of the primary
instead of their own, for which they also do not allocate enough memory,
leading to crashes.

Fixes: 1d88ba171942 ("net/mlx5: refactor Tx data path")
Fixes: 21c8bb4928c9 ("net/mlx5: split Tx queue structure")

Signed-off-by: Olivier Gournet <ogournet@corp.free.fr>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>

net/fm10k: fix VF Tx queue initialization

The SGLORT (Source Global Resource Tag) of the VF queues should be
assigned by PF driver, VF driver should not set these registers.

Fixes: 3b845bf6bcd9 ("fm10k: fix switch manager high CPU usage")

Reported-by: Sarath Somasekharan <sarathx.somasekharan@intel.com>
Signed-off-by: Xiao Wang <xiao.w.wang@intel.com>

net/mlx5: fix link status report

This commit fixes link status report on device start up when
lcs callback is configured.

Fixes: 62072098b54e ("mlx5: support setting link up or down")

Signed-off-by: Olga Shern <olgas@mellanox.com>

net/fm10k: fix out of order Rx read

In vPMD, when we load Rx desc with _mm_loadu_si128, the volatile
modifier will be cast away, allowing the compiler to reorder the
load instructions.

The Rx recv function's correctness is relying on these load
instructions following a strict sequence, reading the descriptors
in reverse order, so we add compiler barrier to prevent compiler
reorder.

Fixes: 7092be8437bd ("fm10k: add vector Rx")

Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>

net/ixgbe: fix out of order Rx read

In vPMD, when we load Rx desc with _mm_loadu_si128, the volatile
modifier will be cast away, allowing the compiler to reorder the
load instructions.

The Rx recv function's correctness is relying on these load
instructions following a strict sequence, reading the descriptors
in reverse order, so we add compiler barrier to prevent compiler
reorder.

Fixes: c95584dc2b18 ("ixgbe: new vectorized functions for Rx/Tx")

Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>

net/i40e: fix out of order Rx read

In vPMD, when we load Rx desc with _mm_loadu_si128, the volatile
modifier will be cast away, allowing the compiler to reorder the
load instructions.

The Rx recv function's correctness is relying on these load
instructions following a strict sequence, reading the descriptors
in reverse order, so we add compiler barrier to prevent compiler
reorder.

Fixes: 9ed94e5bb04e ("i40e: add vector Rx")

Signed-off-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>

net/enic: fix UDP port in flow director

This fixes issue found by Coverity where a typo caused the flow director
UDP IPv4 source port mask to be assigned the destination port mask
supplied by the caller.

Coverity issue: 137860
Fixes: dfbd6a9cb504 ("net/enic: extend flow director support for 1300 series")

Signed-off-by: John Daley <johndale@cisco.com>

net/enic: fix multi-queue Rx performance

The rx_free_thresh was not being initialized and left at 0
on 1/2 of the RQs which could lead to poor multi-queue
performance.

Fixes: 856d7ba7ed22 ("net/enic: support scattered Rx")

Signed-off-by: John Daley <johndale@cisco.com>
Reviewed-by: Nelson Escobar <neescoba@cisco.com>

net/enic: rename functions for queue index conversion

The function names for converting between RQ indexes known to
the RTE code and internal RQ indexes for primary Start of Packet
(SOP) queues and spill-over (Data) queues was unclear and
confusing.

Clarify with more explicit function names.

Signed-off-by: John Daley <johndale@cisco.com>
Reviewed-by: Nelson Escobar <neescoba@cisco.com>

net/enic: fix crash on MTU update or Rx queue reconfigure

The incorrect completion queue corresponding to an RQ would be
freed if multiple Rx queues are in use and the MTU is changed,
or an Rx queue is released. This could lead to a segmentation fault
when the device is disabled or even in the Rx or Tx paths.

The index of the completion queue corresponding to a RQ needed
to be adjusted after Rx scatter was introduced.

Fixes: 856d7ba7ed22 ("net/enic: support scattered Rx")

Signed-off-by: John Daley <johndale@cisco.com>
Reviewed-by: Nelson Escobar <neescoba@cisco.com>

net/enic: revert truncated packets counter fix

The reason this commit was needed was because of a misconfiguration of
the receive queue when not using Rx scatter.  This patch is
unnecessary if the receive queue is configured correctly.

Fixes: d142e1ac1089 ("net/enic: fix calculation of truncated packets")

Signed-off-by: Nelson Escobar <neescoba@cisco.com>
Reviewed-by: John Daley <johndale@cisco.com>

net/enic: fix Rx queue index when not using Rx scatter

The Rx scatter patch was accidentally setting the index of the
secondary receive queue in the primary receive queue's initialization
when the secondary receive queue wasn't needed and was disabled.  This
caused some misleading hardware counters in some situations.

Fixes: 856d7ba7ed22 ("net/enic: support scattered Rx")

Signed-off-by: Nelson Escobar <neescoba@cisco.com>
Reviewed-by: John Daley <johndale@cisco.com>

app/testpmd: fix flow director endianness

The vlan mask and tunnel id mask of flow director are defined as big
endian. So they should be converted.
When the mask is printed, the parameters are not converted either.
This patch converts the mask parameters.

Fixes: 7c554b4f0484 ("app/testpmd: update display of flow director information")
Fixes: 53b2bb9b7ea7 ("app/testpmd: new flow director commands")

Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

app/testpmd: fix flow director mask

In mac-vlan mode, MAC address mask is not supported by HW.
The MAC address mask should not be set in mac-vlan mode.
Remove this parameter from the CLI. Remove MAC address
from mask print too.

Fixes: 53b2bb9b7ea7 ("app/testpmd: new flow director commands")

Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

net/ixgbe: fix flow director mask

In mac-vlan mode, MAC address mask is not supported by HW.
The MAC address mask should not be set in mac-vlan mode.
Instead, only set it in tunnel mode.

Fixes: 82fb702077f6 ("ixgbe: support new flow director modes for X550")

Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>

net/mlx5: fix hash key size retrieval

Return RSS key size in struct rte_eth_dev_info.

Fixes: 0f6f219e7919 ("app/testpmd: fix RSS hash key size")

Signed-off-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>

net/bnx2x: fix build with icc

When bnx2x debug config enabled, icc (ICC) 17.0.0 20160721
gives the following compilation error:

.../drivers/net/bnx2x/elink.c(6682):
error #3656: variable "fw_ver1" may be used before its value is set
PMD_DRV_LOG(DEBUG,
^

According logic in the code, this error is a false positive,
but since this is not in the fast path, fixing compiler error by
assigning initial value to variable.

Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Rasesh Mody <rasesh.mody@qlogic.com>

net/i40e: make vector driver filenames consistent

To be consistent with the naming for ARM NEON implementation,
i40e_rxtx_vec.c is renamed to i40e_rxtx_vec_sse.c.

Signed-off-by: Jianbo Liu <jianbo.liu@linaro.org>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>

net/i40e: implement vector PMD for ARM

Use ARM NEON intrinsic to implement i40e vPMD

Signed-off-by: Jianbo Liu <jianbo.liu@linaro.org>
Acked-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>

net/i40e: extract non-x86 specific code from vector driver

move scalar code which does not use x86 intrinsic functions to new file
"i40e_rxtx_vec_common.h", while keeping x86 code in i40e_rxtx_vec.c.
This allows the scalar code to to be shared among vector drivers for
different platforms.

Signed-off-by: Jianbo Liu <jianbo.liu@linaro.org>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>

vhost: support indirect descriptor in non-mergeable Rx

Linux virtio-net kernel driver uses indirect descriptors when
mergeable buffers are not used.

This patch adds its support, fixing the use of indirect
descriptors with these guests.

Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>

vhost: support indirect descriptor in mergeable Rx

Windows virtio-net driver uses indirect descriptors with
mergeable buffers.

This patch adds its support, fixing the use of indirect
descriptors with these guests.

Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>

vhost: fix use after free

Fix the coverity USE_AFTER_FREE issue.

Coverity issue: 137884
Fixes: a277c7159876 ("vhost: refactor code structure")

Reported-by: John McNamara <john.mcnamara@intel.com>
Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>

vhost: retrieve available head once

There is no need to retrieve the latest avail head every time we enqueue
a packet in the mereable Rx path by

    avail_idx = *((volatile uint16_t *)&vq->avail->idx);

Instead, we could just retrieve it once at the beginning of the enqueue
path. This could diminish the cache penalty slightly, because the virtio
driver could be updating it while vhost is reading it (for each packet).

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Signed-off-by: Zhihong Wang <zhihong.wang@intel.com>
Reviewed-by: Jianbo Liu <jianbo.liu@linaro.org>
Tested-by: Maxime Coquelin <maxime.coquelin@redhat.com>

vhost: prefetch available ring

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Signed-off-by: Zhihong Wang <zhihong.wang@intel.com>
Reviewed-by: Jianbo Liu <jianbo.liu@linaro.org>
Tested-by: Maxime Coquelin <maxime.coquelin@redhat.com>

vhost: shadow used ring update

The basic idea is to shadow the used ring update: update them into a
local buffer first, and then flush them all to the virtio used vring
at once in the end.

And since we do avail ring reservation before enqueuing data, we would
know which and how many descs will be used. Which means we could update
the shadow used ring at the reservation time. It also introduce another
slight advantage: we don't need access the desc->flag any more inside
copy_mbuf_to_desc_mergeable().

Signed-off-by: Zhihong Wang <zhihong.wang@intel.com>
Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Reviewed-by: Jianbo Liu <jianbo.liu@linaro.org>
Tested-by: Maxime Coquelin <maxime.coquelin@redhat.com>

vhost: use last available index for ring reservation

shadow_used_ring will be introduced later. Since then last avail
idx will not be updated together with last used idx.

So, here we use last_avail_idx for avail ring reservation.

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Signed-off-by: Zhihong Wang <zhihong.wang@intel.com>
Reviewed-by: Jianbo Liu <jianbo.liu@linaro.org>
Tested-by: Maxime Coquelin <maxime.coquelin@redhat.com>

vhost: simplify mergeable Rx vring reservation

Let it return "num_buffers" we reserved, so that we could re-use it
with copy_mbuf_to_desc_mergeable() directly, instead of calculating
it again there.

Meanwhile, the return type of copy_mbuf_to_desc_mergeable is changed
to "int". -1 will be return on error.

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Signed-off-by: Zhihong Wang <zhihong.wang@intel.com>
Reviewed-by: Jianbo Liu <jianbo.liu@linaro.org>
Tested-by: Maxime Coquelin <maxime.coquelin@redhat.com>

vhost: optimize cache access

This patch reorders the code to delay virtio header write to improve
cache access efficiency for cases where the mrg_rxbuf feature is turned
on. CPU pipeline stall cycles can be significantly reduced.

Virtio header write and mbuf data copy are all remote store operations
which takes a long time to finish. It's a good idea to put them together
to remove bubbles in between, to let as many remote store instructions
as possible go into store buffer at the same time to hide latency, and
to let the H/W prefetcher goes to work as early as possible.

On a Haswell machine, about 100 cycles can be saved per packet by this
patch alone. Taking 64B packets traffic for example, this means about 60%
efficiency improvement for the enqueue operation.

Signed-off-by: Zhihong Wang <zhihong.wang@intel.com>
Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Reviewed-by: Jianbo Liu <jianbo.liu@linaro.org>
Tested-by: Maxime Coquelin <maxime.coquelin@redhat.com>

vhost: remove useless volatile

last_used_idx is a local var, there is no need to decorate it
by "volatile".

Signed-off-by: Zhihong Wang <zhihong.wang@intel.com>
Reviewed-by: Jianbo Liu <jianbo.liu@linaro.org>
Tested-by: Maxime Coquelin <maxime.coquelin@redhat.com>

vhost: access header only if offloading is supported

If offloading features are not negotiated, parsing the virtio header
is not needed.

Micro-benchmark with testpmd shows that the gain is +4% with indirect
descriptors, +1% when using direct descriptors.

Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Acked-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>

ethdev: prevent duplicate event callback

This change prevents the attempt to add a structure which is
already on the callback list. If a struct with matching
parameters is found on the list, then no action is taken.

Fixes: ac2f69c ("ethdev: fix crash if malloc of user callback fails")

Signed-off-by: E. Scott Daniels <daniels@research.att.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>

mempool: fix search of maximum contiguous pages

paddr[i] + pg_sz always points to the start physical address of the
2nd page after pddr[i], so only up to 2 pages can be combinded to
be used. With this revision, more than 2 pages can be used.

Fixes: 84121f197187 ("mempool: store memory chunks in a list")

Signed-off-by: Wei Dai <wei.dai@intel.com>
Signed-off-by: Olivier Matz <olivier.matz@6wind.com>

app/testpmd: fix PF/VF check of flow director

Parameters pf & vf are added into most of flow director
filter CLIs.
But mac-vlan and tunnel filters don't have these parameters,
the parameters should not be checked for mac-vlan and tunnel
filters.

Fixes: e6a68c013353 ("app/testpmd: extend commands for flow director in VF")

Reported-by: Frederico Cadete <frederico.cadete-ext@oneaccess-net.com>
Signed-off-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

crypto: register aliases for renamed vdev drivers

This registers the legacy names of the driver being renamed in
commit 2f45703c17ac ("drivers: make driver names consistent").

Signed-off-by: Jan Blunck <jblunck@infradead.org>
Tested-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Reviewed-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>

net: register aliases for renamed vdev drivers

This registers the legacy names of the driver being renamed in
commit 2f45703c17ac ("drivers: make driver names consistent").

Signed-off-by: Jan Blunck <jblunck@infradead.org>
Reviewed-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>

eal: add driver name alias

This adds infrastructure for drivers to allow being requested by an alias
so that a renamed driver can still get loaded by its legacy name.

Signed-off-by: Jan Blunck <jblunck@infradead.org>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Tested-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Reviewed-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>

examples/l3fwd: use accelerated CRC on ARM

if machine level CRC extension are available, offload the
hash to machine provide functions e.g. armv8-a CRC extensions
support it

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Reviewed-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>

examples/tep_term: fix packet length with multi-segments

For multi-seg mbuf, ip->total_length should be pkt_len subtract
ether len.

Fixes: 4abe471ed6fc ("examples/tep_term: implement VXLAN processing")

Signed-off-by: Michael Qiu <qiudayu@chinac.com>
Acked-by: Jianfeng Tan <jianfeng.tan@intel.com>

examples/tep_term: fix L4 length

l4_len is not fixed, althrough mostly it is a fixed value,
but when guest using iperf to do some tests, the l4_len
will have another 12 bytes optional fields.

Fixes: 2bb43bd4350a ("examples/tep_term: add TSO offload configuration")

Signed-off-by: Michael Qiu <qiudayu@chinac.com>
Acked-by: Jianfeng Tan <jianfeng.tan@intel.com>

examples/ip_pipeline: fix build with gcc 6

This patch fixes the misleading indentation error on compiling
ip_pipeline app with gcc v6.2.1.

Fixes: 3f2c9f3bb6c6 ("examples/ip_pipeline: add TAP port")

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

examples/ip_pipeline: fix build on BSD

Error log:
 CC init.o
 examples/ip_pipeline/init.c:38:22: fatal error: linux/if.h: No such file or directory
 #include <linux/if.h>

Fixes: 3f2c9f3bb6c6 ("examples/ip_pipeline: add TAP port")

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

kni: fix build with kernel 4.9

compile error:
  CC [M]  .../lib/librte_eal/linuxapp/kni/igb_main.o
.../lib/librte_eal/linuxapp/kni/igb_main.c:2317:21:
error: initialization from incompatible pointer type
[-Werror=incompatible-pointer-types]
  .ndo_set_vf_vlan = igb_ndo_set_vf_vlan,
                     ^~~~~~~~~~~~~~~~~~~

Linux kernel 4.9 updates API for ndo_set_vf_vlan:
Linux: 79aab093a0b5 ("net: Update API for VF vlan protocol 802.1ad support")

Use new API for Linux kernels >= 4.9

Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Tested-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

kni: fix build with kernel < 3.1

compile error:
  CC [M]  .../lib/librte_eal/linuxapp/kni/kni_misc.o
cc1: warnings being treated as errors
.../lib/librte_eal/linuxapp/kni/kni_misc.c: In function ‘kni_exit_net’:
.../lib/librte_eal/linuxapp/kni/kni_misc.c:113:18:
error: unused variable ‘knet’

For kernel versions < v3.1 mutex_destroy() is a macro and does nothing,
this cause an unused variable warning for knet which used in the
mutex_destroy()

mutex_destroy() converted into static inline function with commit:
Linux: 4582c0a4866e ("mutex: Make mutex_destroy() an inline function")

To fix the warning unused attribute added to the knet variable.

Fixes: 93a298b34e1b ("kni: support core id parameter in single threaded mode")

Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>

version: 16.11-rc1

Signed-off-by: Thomas Monjalon <thomas.monjalon@6wind.com>

doc: fix libcrypto guide title

Libcrypto documentation was missing the equal signs ("="),
in its title, so it was not present in the documentation
generated.

Fixes: d61f70b4c918 ("crypto/libcrypto: add driver for OpenSSL library")

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

doc: how to build KASUMI as shared library

Libsso KASUMI library has to be built with specific
parameters to make the KASUMI PMD be built as
a shared library, so a note has been added in
its documentation.

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

doc: add ZUC shared library limitation

ZUC PMD cannot be built as a shared library, due to
the fact that some assembly code in the underlying libsso
library is not relocatable.
This will be fixed in the future, but for the moment,
it is added as a limitationof the PMD.

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

crypto/libcrypto: fix init function names

All init/uninit function names in the virtual devices have been renamed,
so they finish with probe or remove, so to keep consistency,
same thing should be done in this PMD.

Fixes: d61f70b4c918 ("crypto/libcrypto: add driver for OpenSSL library")

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Acked-by: Fiona Trahe <fiona.trahe@intel.com>

crypto/zuc: fix init function names

All init/uninit function names in the virtual devices have been renamed,
so they finish with probe or remove, so to keep consistency,
same thing should be done in this PMD.

Fixes: cf7685d68f00 ("crypto/zuc: add driver for ZUC library")

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Acked-by: Fiona Trahe <fiona.trahe@intel.com>

app/test: store digest/IV at a DMA able address

For physical crypto devices, IV and digest are processed by the crypto
device which needs the contents to be written on some DMA able address.

So in order to do that, IV and digest are accomodated in the packet.

Signed-off-by: Akhil Goyal <akhil.goyal@nxp.com>
Acked-by: Fiona Trahe <fiona.trahe@intel.com>

app/test: fix crypto mbuf pool size

The created pool for crypto mbufs may be too big in some
environments.
To avoid this issue, mbuf pool is reverted to its previous size.
Moreover, here is added additional small pool with one large mbuf,
that is needed in large data test scenarios.

Fixes: ffbe3be0d4b5 ("app/test: add libcrypto")

Signed-off-by: Piotr Azarewicz <piotrx.t.azarewicz@intel.com>
Acked-by: Michal Jastrzebski <michalx.k.jastrzebski@intel.com>

examples/l2fwd-crypto: fix verify with decrypt in chain

This patch fixes crypto operation data parameters setting
in l2fwd-crypto application, making decryption in chain
with auth verification work.

How to reproduce the issue:

1. Run l2fwd_crypto with command:
-c 0x3 -n 4 --vdev "crypto_aesni_mb" \
--vdev "crypto_aesni_mb" \
-- -p 0x3 --chain CIPHER_HASH \
--cipher_op ENCRYPT --cipher_algo AES_CBC \
--cipher_key 00:01:02:03:04:05:06:07:08:09:0a:0b:0c:0d:0e:0f \
--iv 00:01:02:03:04:05:06:07:08:09:0a:0b:0c:0d:0e:ff \
--auth_op GENERATE --auth_algo SHA1_HMAC \
--auth_key
11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:
11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:
11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11

2. Send packet with payload and capture forwarded packet.
Payload in forwarded packet is encrypted, what is good.

3. Run l2fwd_crypto with command:
-c 0x3 -n 4 --vdev "crypto_aesni_mb" \
--vdev "crypto_aesni_mb" \
-- -p 0x3 --chain HASH_CIPHER \
--cipher_op DECRYPT --cipher_algo AES_CBC \
--cipher_key 00:01:02:03:04:05:06:07:08:09:0a:0b:0c:0d:0e:0f \
--iv 00:01:02:03:04:05:06:07:08:09:0a:0b:0c:0d:0e:ff \
--auth_op VERIFY --auth_algo SHA1_HMAC \
--auth_key
11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:
11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:
11:11:11:11:11:11:11:11:11:11:11:11:11:11:11:11

4. Send earlier captured packet and capture forwarded packet.
Payload in newly captured packet is not decrypted, what is wrong.

Fixes: 387259bd6c67 ("examples/l2fwd-crypto: add sample application")

Signed-off-by: Piotr Azarewicz <piotrx.t.azarewicz@intel.com>
Acked-by: Michal Jastrzebski <michalx.k.jastrzebski@intel.com>

app/test: reduce libcrypto operations number

This patch reduce the number of total operations from 1M to 10K,
because test is taking too long time now.

Fixes: ffbe3be0d4b5 ("app/test: add libcrypto")

Signed-off-by: Marcin Kerlin <marcinx.kerlin@intel.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

net/ixgbe: add application callback on VF to PF messsage

call _rte_eth_dev_callback_process from ixgbe_rcv_msg_from_vf function.

The callback asks the user application if it is allowed to perform
the function.

If the cb_param.retval is RTE_PMD_IXGBE_MB_EVENT_PROCEED then continue,
if 0, do nothing and send ACK to VF
if > 1, do nothing and send NAK to VF.

Signed-off-by: Alex Zelezniak <alexz@att.com>
Signed-off-by: Bernard Iremonger <bernard.iremonger@intel.com>

ethdev: add an argument to internal callback function

add cb_arg parameter to the _rte_eth_dev_callback_process function.

Adding a parameter to this function allows passing information
to the application when an eth device event occurs such as
a VF to PF message.
This allows the application to decide if a particular function
is permitted.

Signed-off-by: Bernard Iremonger <bernard.iremonger@intel.com>
Signed-off-by: Alex Zelezniak <alexz@att.com>

drivers: rename register macro prefix

All macros related to driver registeration renamed from DRIVER_*
to RTE_PMD_*

This includes:

 DRIVER_REGISTER_PCI -> RTE_PMD_REGISTER_PCI
 DRIVER_REGISTER_PCI_TABLE -> RTE_PMD_REGISTER_PCI_TABLE
 DRIVER_REGISTER_VDEV -> RTE_PMD_REGISTER_VDEV
 DRIVER_REGISTER_PARAM_STRING -> RTE_PMD_REGISTER_PARAM_STRING
 DRIVER_EXPORT_* -> RTE_PMD_EXPORT_*

Fix PMDINFOGEN tool to look for matches of RTE_PMD_REGISTER_*.

Signed-off-by: Shreyansh Jain <shreyansh.jain@nxp.com>
Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>

net/i40e: implement new Rx checksum flag

Add CKSUM_GOOD flag to distinguish a good checksum from an unknown one.

Signed-off-by: Xiao Wang <xiao.w.wang@intel.com>
Acked-by: Jing Chen <jing.d.chen@intel.com>

net/ixgbe: implement new Rx checksum flag

Add CKSUM_GOOD flag to distinguish a good checksum from an unknown one.

Signed-off-by: Xiao Wang <xiao.w.wang@intel.com>
Acked-by: Jing Chen <jing.d.chen@intel.com>

net/e1000: implement new Rx checksum flag

Add CKSUM_GOOD flag to distinguish a good checksum from an unknown one.

Signed-off-by: Xiao Wang <xiao.w.wang@intel.com>
Acked-by: Jing Chen <jing.d.chen@intel.com>

net/fm10k: implement new Rx checksum flag

Add CKSUM_GOOD flag to distinguish a good checksum from an unknown one.
Since vector Rx supports checksum offload, this patch removes the
hw_ip_checksum check in fm10k_rx_vec_condition_check().

Signed-off-by: Xiao Wang <xiao.w.wang@intel.com>
Acked-by: Jing Chen <jing.d.chen@intel.com>

net/fm10k: fix Rx checksum flags

A previous patch removed some necessary lines about Rx checksum offload
by mistake, this patch adds them back.

Fixes: 6046898f5097 ("net/mbuf: remove unused Rx error flags")

Signed-off-by: Xiao Wang <xiao.w.wang@intel.com>
Acked-by: Jing Chen <jing.d.chen@intel.com>

net/ixgbe: support checksum flags in SSE vector Rx

Update desc_to_olflags_v() to set PKT_RX_IP_CKSUM_BAD and
PKT_RX_L4_CKSUM_BAD in the ol_fags of the mbuf.

The Rx vector function can now be used with hw_ip_checksum
enabled.

Tested with:

  cd dpdk.org/
  make config T=x86_64-native-linuxapp-gcc
  make -j32
  mkdir -p /mnt/huge
  mount -t hugetlbfs nodev /mnt/huge
  echo 256 > /sys/devices/system/node/node0/hugepages/hugepages-2048kB/nr_hugepages
  modprobe uio_pci_generic
  python tools/dpdk_nic_bind.py -b uio_pci_generic 0000:04:00.0
  ./build/app/testpmd -l 2,4 -- --total-num-mbufs=65536 -i --port-topology=chained --enable-rx-cksum --disable-hw-vlan-filter --disable-hw-vlan-strip
    set fwd rxonly
    set verbose 1
    start

  # send packets to testpmd using scapy
  eh = Ether(src="00:01:02:03:04:05", dst="00:1B:21:AB:8F:10")
  p = Ether()/IP(src="1.1.1.1", dst="1.1.1.2")/UDP()/Raw("x"*50)
  sendp(p, iface="ixgbe2")
  p = Ether()/IP(src="1.1.1.1", dst="1.1.1.2", chksum=42)/UDP()/Raw("x"*50)
  sendp(p, iface="ixgbe2")
  p = Ether()/IP(src="1.1.1.1", dst="1.1.1.2")/UDP(chksum=42)/Raw("x"*50)
  sendp(p, iface="ixgbe2")
  p = Ether()/IP(src="1.1.1.1", dst="1.1.1.2", chksum=45)/UDP(chksum=42)/Raw("x"*50)
  sendp(p, iface="ixgbe2")

  # result
  port 0/queue 0: received 1 packets
    src=00:01:02:03:04:05 - dst=00:1B:21:AB:8F:10 - type=0x0800 - length=92 - nb_segs=1Unknown packet type
   - Receive queue=0x0
  port 0/queue 0: received 1 packets
    src=00:01:02:03:04:05 - dst=00:1B:21:AB:8F:10 - type=0x0800 - length=92 - nb_segs=1Unknown packet type
   - Receive queue=0x0
    PKT_RX_IP_CKSUM_BAD
  port 0/queue 0: received 1 packets
    src=00:01:02:03:04:05 - dst=00:1B:21:AB:8F:10 - type=0x0800 - length=92 - nb_segs=1Unknown packet type
   - Receive queue=0x0
    PKT_RX_L4_CKSUM_BAD
  port 0/queue 0: received 1 packets
    src=00:01:02:03:04:05 - dst=00:1B:21:AB:8F:10 - type=0x0800 - length=92 - nb_segs=1Unknown packet type
   - Receive queue=0x0
    PKT_RX_L4_CKSUM_BAD
    PKT_RX_IP_CKSUM_BAD

Signed-off-by: Maxime Leroy <maxime.leroy@6wind.com>
Signed-off-by: Olivier Matz <olivier.matz@6wind.com>
Acked-by: Remy Horton <remy.horton@intel.com>

app/testpmd: add ixgbe VF management

add test for set vf vlan anti spoof
add test for set vf mac anti spoof
add test for set vf vlan stripq
add test for set vf vlan insert
add test for set tx loopback
add test for set all queues drop enable bit
add test for set vf split drop enable bit
add test for set vf mac address
add new APIs to testpmd help command.

add new APIs to the testpmd guide

Modify Makefile, add library rte_pmd_ixgbe to _LDLIBS-y.

Signed-off-by: Bernard Iremonger <bernard.iremonger@intel.com>

net/ixgbe: add API for VF management

Add APIs to configure and manage VFs on an Intel 82559 NIC.

add rte_pmd_ixgbe_set_vf_vlan_anti_spoof function.
add rte_pmd_ixgbe_set_vf_mac_anti_spoof function.
add rte_pmd_ixgbe_set_vf_stripq function.

Signed-off-by: Alex Zelezniak <alexz@att.com>
add rte_pmd_ixgbe_set_vf_vlan_insert function.
add rte_pmd_ixgbe_set_tx_loopback function.
add rte_pmd_ixgbe_set_all_queues_drop function.
add rte_pmd_ixgbe_set_vf_split_drop_en function.
add rte_pmd_ixgbe_set_vf_mac_addr function.

Signed-off-by: Bernard Iremonger <bernard.iremonger@intel.com>

net/ring: fix parameter string info

Fixes: 65eca099f405 ("drivers: split parameters infos in multiple lines")

Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>

net/ring: fix ring device creation via devargs

Using nodeaction devarg lets creating multiple ring eth devices:
"eth_ring0,nodeaction=R0:0:CREATE,nodeaction=R1:0:CREATE"

Trying to create all devices with same name fails. Since first part of
the nodeaction devarg is name (in above sample R0,R1), this name field
can be used as eth dev name.

Fixes: 61934c0956d4 ("ring: convert to use of PMD_REGISTER_DRIVER and fix linking")

Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>

net/bnxt: support async link notification

This patch adds support to get Link notification asynchronously.
The HW sends async notifications on default completion ring. The
PMD processes these notifications and logs a message appropriately.

Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

doc: update bnxt guide

This patch reformats the Broadcom PMD driver documentation.

Also since the PMD now loads on a VF interface, update the
documentation accordingly.

Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Acked-by: John McNamara <john.mcnamara@intel.com>

net/bnxt: fix crash when closing

This patch fixes segfault encountered during dev_uninit/close routine.
KNI sample app can be used to reproduce the issue.

Fixes: c09f57b49c13 ("net/bnxt: add start/stop/link update operations")

Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/bnx2x: merge debug register operations into headers

The register read/writes should just be static inline instead of
alternately defined as routines or macros depending on the status of
debugging.

Fix bnx2x_reg_read32() returning 0 during debug unaligned reads.

Fixes: b5bf7719221d ("bnx2x: driver support routines")

Signed-off-by: Chas Williams <3chas3@gmail.com>
Acked-by: Harish Patil <harish.patil@qlogic.com>

net/bnx2x: do not return structs

bnx2x_loop_obtain_resources() returns a struct.  This routine either
succeeds or fails -- We don't need a struct for that.

Fixes: 540a211084a7 ("bnx2x: driver core")

Signed-off-by: Chas Williams <3chas3@gmail.com>
Acked-by: Harish Patil <harish.patil@qlogic.com>

net/bnx2x: check return codes during VF mailbox operation

Refactor bnx2x_do_req4pf() to be easier to read and return errors when
the transaction fails -- Previously, it could succeed when the control
channel was down.

Fixes: 540a211084a7 ("bnx2x: driver core")

Signed-off-by: Chas Williams <3chas3@gmail.com>
Acked-by: Harish Patil <harish.patil@qlogic.com>

net/bnx2x: serialize access to VF mailbox

The pf2vf mailbox can only be used by one thread at a time.

Fixes: 540a211084a7 ("bnx2x: driver core")

Signed-off-by: Chas Williams <3chas3@gmail.com>
Acked-by: Harish Patil <harish.patil@qlogic.com>

net/bnx2x: replace macro with static function

Replace BNX2X_TLV_APPEND() with the clearer and safer bnx2x_add_tlv().
bnx2x_add_tlv() was previously prototyped at some point but can be static.

Fixes: 540a211084a7 ("bnx2x: driver core")

Signed-off-by: Chas Williams <3chas3@gmail.com>
Acked-by: Harish Patil <harish.patil@qlogic.com>

net/bnx2x: restrict Rx mask flags sent to the PF

Don't use bnx2x_fill_accept_flags() to fill the RX mask in the VF
since the PF only handles a subset of the existing flags.  now,
bnx2x_fill_accept_flags() can be static.

Fixes: 540a211084a7 ("bnx2x: driver core")

Signed-off-by: Chas Williams <3chas3@gmail.com>
Acked-by: Harish Patil <harish.patil@qlogic.com>

net/bnx2x: remove unused Rx queue code

Fixes: 540a211084a7 ("bnx2x: driver core")

Signed-off-by: Chas Williams <3chas3@gmail.com>
Acked-by: Harish Patil <harish.patil@qlogic.com>

net/bnx2x: remove delay during device startup

This 2.5s delay doesn't seem to serve any purpose other than a being a
pause after logging the device configuration.

Fixes: 540a211084a7 ("bnx2x: driver core")

Signed-off-by: Chas Williams <3chas3@gmail.com>
Acked-by: Harish Patil <harish.patil@qlogic.com>

net/bnx2x: remove unused preprocessor code

ELINK_INCLUDE_EMUL and ELINK_INCLUDE_FPGA are never defined.  Remove them
along with enumeration constants dependent on their inclusion.

Fixes: 540a211084a7 ("bnx2x: driver core")

Signed-off-by: Chas Williams <3chas3@gmail.com>
Acked-by: Harish Patil <harish.patil@qlogic.com>

net/bnx2x: get cache line size from build configuration

Correctly hint the cache line size.  Remove unused macros associated
with the cache line size.

Fixes: 540a211084a7 ("bnx2x: driver core")

Signed-off-by: Chas Williams <3chas3@gmail.com>
Acked-by: Harish Patil <harish.patil@qlogic.com>

net/mlx5: fix Rx function selection

mlx5_rx_queue_setup() was setting the Rx function by itself instead of
using priv_select_rx_function() written for that purpose.

Fixes: cdab90cb5c8d ("net/mlx5: add Tx/Rx burst function selection wrapper")

Signed-off-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>

net/enic: add ethernet VLAN packet type

Enic is capable of recognizing packets to be delivered to the
app with single VLAN tags. Advertise this with the ptype
RTE_PTYPE_L2_ETHER_VLAN and set the ptype for VLAN packets.

Signed-off-by: John Daley <johndale@cisco.com>

doc: add ixgbe supported chipsets and NICs

get the list of all chipsets and NICs supported
by ixgbe driver from
https://downloadcenter.intel.com/download/14687

Signed-off-by: Wei Dai <wei.dai@intel.com>
Acked-by: John McNamara <john.mcnamara@intel.com>

net/bnxt: fix bit shift size

Some(?) compilers will treat the unmarked constant 1 as a 32-bit
integer, but the shift operation is in a loop that could run up to
63 times -- undefined behavior!

Coverity issue: 127546
Fixes: 778b759ba10e ("net/bnxt: add MAC address")

Signed-off-by: John W. Linville <linville@tuxdriver.com>
Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>