net/virtio: fix incorrect cast of void *

The rx_queues and tx_queues fields of the data structure points to a struct
virtnet_rx or virtnet_tx. Casting it to a virtqueue is an error.

It does not trigger any bug because pointer is not dereferenced inside the
function, but it can become a bug if this code is copy/pasted and vq is
dereferenced.

Fixes: 01ad44fd374f ("net/virtio: split Rx/Tx queue")
Cc: stable@dpdk.org
Signed-off-by: Didier Pallard <didier.pallard@6wind.com>
Signed-off-by: Olivier Matz <olivier.matz@6wind.com>

app/testpmd: dump more port info

Display some more useful information from dev_info when
calling "show port info <portid>".

Signed-off-by: Olivier Matz <olivier.matz@6wind.com>

net/virtio: fix Rx and Tx handler selection for arm32

On arm32, we were always selecting the simple handler, but it is only
available if neon is present.

This is due to a typo in the name of the config option.
CONFIG_RTE_ARCH_ARM is for Makefiles. One should use RTE_ARCH_ARM.

Fixes: 2d7c37194ee4 ("net/virtio: add NEON based Rx handler")
Cc: stable@dpdk.org
Signed-off-by: Samuel Gauthier <samuel.gauthier@6wind.com>

net/bntx: use new API to get iova address XXX

XXX: use mbuf instead of m ?

The functions rte_mbuf_data_dma_addr*() are deprecated.
Replace them with their iova version. Also rename the 'data' variable
to 'm' (for mbuf) to avoid confusion.

Signed-off-by: Olivier Matz <olivier.matz@6wind.com>

net/bnxt: fix headroom initialization

When allocating a new mbuf for Rx, the value of m->data_off should not
be reset to its default value (RTE_PKTMBUF_HEADROOM), instead of reusing
the previous undefined value, which could cause the packet to have a too
small or too high headroom.

Fixes: 2eb53b134aae ("net/bnxt: add initial Rx code")
Cc: stable@dpdk.org
Signed-off-by: Zijie Pan <zijie.pan@6wind.com>
Signed-off-by: Didier Pallard <didier.pallard@6wind.com>
Signed-off-by: Olivier Matz <olivier.matz@6wind.com>

mk: fix libs installation when installing sdk

The 'install-sdk' target creates an invalid symlink in the install
directory:

  # make install-sdk DESTDIR=/tmp/toto V=1
  [...]
  ln -snf $(/root/dpdk.org/buildtools/relpath.sh /tmp/toto/usr/local/lib \
            /tmp/toto/usr/local/share/dpdk/x86_64-native-linuxapp-gcc/) \
       /tmp/toto/usr/local/share/dpdk/x86_64-native-linuxapp-gcc/lib
  [...]
  # ls -l /tmp/toto/usr/local/share/dpdk/x86_64-native-linuxapp-gcc
  drwxr-xr-x 2 root root 4096 Dec 11 11:49 app
  lrwxrwxrwx 1 root root   21 Dec 11 11:49 include -> ../../../include/dpdk
  lrwxrwxrwx 1 root root   12 Dec 11 11:49 lib -> ../../../lib     # WRONG LINK

This happens because the directory is not created. Moreover, it makes
sense for install-sdk to install the libs, which could be necessary to
build something against dpdk. After the patch, the link is correct and
the *.a libs are properly installed:

  # ls -l /tmp/toto/usr/local/share/dpdk/x86_64-native-linuxapp-gcc
  drwxr-xr-x 2 root root 4096 Dec 11 11:53 app
  lrwxrwxrwx 1 root root   21 Dec 11 11:53 include -> ../../../include/dpdk
  lrwxrwxrwx 1 root root   12 Dec 11 11:53 lib -> ../../../lib
  # ls -l /tmp/toto/usr/local/share/dpdk/x86_64-native-linuxapp-gcc/lib/
  -rw-r--r-- 1 root root   1378 Dec  8 14:08 libdpdk.a
  -rw-r--r-- 1 root root 123380 Dec  8 14:02 librte_acl.a
  -rw-r--r-- 1 root root   3406 Dec  8 14:02 librte_bitratestats.a
  [...]

Fixes: 6efca1e9f873 ("mk: split install rule")
Cc: stable@dpdk.org
Signed-off-by: Samuel Gauthier <samuel.gauthier@6wind.com>
Signed-off-by: Olivier Matz <olivier.matz@6wind.com>

ethdev: fix missing imissed counter in xstats

imissed counter has been set as deprecated in commit 49f386542af4
("ethdev: remove driver specific stats") and removed from the
rte_eth_xstats_name_off structure.

The imissed counter has been restored few commits later but has not been
restored in the rte_eth_stats structure. Add it back again.

Fixes: 4eadb8ba11b7 ("ethdev: do not deprecate imissed counter")
Cc: stable@dpdk.org
Signed-off-by: Thibaut Collet <thibaut.collet@6wind.com>
Signed-off-by: Olivier Matz <olivier.matz@6wind.com>

scripts: enhance the get maintainer error message

When the DPDK_GETMAINTAINER_PATH variable was not set, the error
message from get-maintainer.sh was quite cheap:

$ devtools/get-maintainer.sh --help
usage: get-maintainer.sh <patch>
Cannot execute DPDK_GETMAINTAINER_PATH

Add a more detailed explanation about this variable in the help.

Signed-off-by: Olivier Matz <olivier.matz@6wind.com>

test: add a testcase for dynamic logs

Signed-off-by: Olivier Matz <olivier.matz@6wind.com>

test: rely on dynamic log level to display hexdumps

Signed-off-by: Olivier Matz <olivier.matz@6wind.com>

app/testpmd: register a specific log type

Instead of using USERx logs, register a dynamic log type, as introduced in
commit c1b5fa94a46f ("eal: support dynamic log types").

Signed-off-by: Olivier Matz <olivier.matz@6wind.com>

mbuf: remove control mbuf

The rte_ctrlmbuf structure is not used by any example application
in dpdk. Remove it, as announced on the mailing list.

Signed-off-by: Olivier Matz <olivier.matz@6wind.com>

mbuf: fix mbuf free performance with non atomic refcnt

When RTE_MBUF_REFCNT_ATOMIC=n, the decrement of the mbuf reference
counter uses an atomic operation. This is not necessary and impacts
the performance (seen with TRex traffic generator).

We cannot replace rte_atomic16_add_return() by rte_mbuf_refcnt_update()
because it would add an additional check.

Solves this by introducing __rte_mbuf_refcnt_update(), which
updates the reference counter without doing anything else.

Fixes: 8f094a9ac5d7 ("mbuf: set mbuf fields while in pool")
Suggested-by: Hanoch Haim <hhaim@cisco.com>
Signed-off-by: Olivier Matz <olivier.matz@6wind.com>

net: update licence for network headers

To be compliant with the DPDK licensing guidelines, switch to
BSD-3-Clause. It can be done safely since the BSD headers from which
these files derive also exist as a BSD-3-Clause license in FreeBSD.

Link: https://raw.githubusercontent.com/freebsd/freebsd/78a6b0861813af31e1354fa407c5701e8764b4d6/sys/netinet/ip_icmp.h
Link: https://raw.githubusercontent.com/freebsd/freebsd/78a6b0861813af31e1354fa407c5701e8764b4d6/sys/netinet/ip.h
Link: https://raw.githubusercontent.com/freebsd/freebsd/78a6b0861813af31e1354fa407c5701e8764b4d6/sys/netinet/sctp.h
Link: https://raw.githubusercontent.com/freebsd/freebsd/78a6b0861813af31e1354fa407c5701e8764b4d6/sys/netinet/tcp.h
Link: https://raw.githubusercontent.com/freebsd/freebsd/78a6b0861813af31e1354fa407c5701e8764b4d6/sys/netinet/udp.h
Signed-off-by: Olivier Matz <olivier.matz@6wind.com>

eal: set affinity for control threads

The management threads must not bother the dataplane or service cores.
Set the affinity of these threads accordingly.

Signed-off-by: Olivier Matz <olivier.matz@6wind.com>

eal: set name when creating a control thread

To avoid code duplication, add a parameter to rte_ctrl_thread_create()
to specify the name of the thread.

This requires to add a wrapper for the thread start routine in
rte_thread_init(), which will first wait that the thread is configured.

Signed-off-by: Olivier Matz <olivier.matz@6wind.com>

eal: new function to create control threads

Many parts of dpdk use their own management threads. Introduce a new
wrapper for thread creation that will be extended in next commit to set
the name and affinity.

To be consistent with other DPDK APIs, the return value is negative in
case of error, which was not the case for pthread_create().

Signed-off-by: Olivier Matz <olivier.matz@6wind.com>

eal: use sizeof to avoid a double use of a define

Only a cosmetic change.

Signed-off-by: Olivier Matz <olivier.matz@6wind.com>

pdump: fix error code check when creating/canceling pthread

On error, pthread_create() returns a positive number (an errno)
but does not set the errno variable.

Fixes: 278f945402c5 ("pdump: add new library for packet capture")
Cc: stable@dpdk.org
Signed-off-by: Olivier Matz <olivier.matz@6wind.com>

vhost: fix error code check when creating pthread

On error, pthread_create() returns a positive number (errno).
Fix the test on the return value.

Fixes: af1475918124 ("vhost: introduce API to start a specific driver")
Fixes: e623e0c6d8a5 ("vhost: add reconnect ability")
Cc: stable@dpdk.org
Signed-off-by: Olivier Matz <olivier.matz@6wind.com>

version: 18.02-rc0

Signed-off-by: Thomas Monjalon <thomas@monjalon.net>

doc: add template release notes for 18.02

Add template release notes for DPDK 18.02 with inline
comments and explanations of the various sections.

Signed-off-by: John McNamara <john.mcnamara@intel.com>

version: 17.11.0

Signed-off-by: Thomas Monjalon <thomas@monjalon.net>

doc: fix GSO guide as PDF

The PDF cannot be built because of some images integration being
forced as SVG. They are converted for PDF format, so the extension
must be a wildcard in the RST file.

Fixes: f6010c7655cc ("doc: add GSO programmer's guide")
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>

net/mlx5: fix flow mark action on port start

Mark action flag was not set on the flow, causing the rx burst function
after port start to ignore the flow mark.

Fixes: 8086cf08b2f0 ("net/mlx5: handle RSS hash configuration in RSS flow")
Signed-off-by: Shahaf Shuler <shahafs@mellanox.com>
Acked-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>

maintainers: move NXP bus drivers in related section

A bus maintainers section has been introduced in 17.11.
Let's move all bus drivers in the same place.

Signed-off-by: Thomas Monjalon <thomas@monjalon.net>

doc: add bus libraries to release notes

Fixes: 919eeaccb2ba ("bus/dpaa: introduce NXP DPAA bus driver skeleton")
Fixes: 7e7df6d0a41d ("bus/fslmc: introduce fsl-mc bus driver")
Fixes: c752998b5e2e ("pci: introduce library and driver")
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>

doc: announce behaviour change to i40e RSS

Add deprecation notice to highlight a future change in the
default configuration behavior for flexible payload for RSS.

Signed-off-by: John McNamara <john.mcnamara@intel.com>
Acked-by: Andrey Chilikin <andrey.chilikin@intel.com>

pdump: fix typo in error log

Fixes: 278f945402c5 ("pdump: add new library for packet capture")
Signed-off-by: Maria Lingemark <maria.lingemark@ericsson.com>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>

mbuf: fix segments number type increase

Update types of variables to correspond to nb_segs type change from
uint8_t to uint16_t.

Fixes: 97cb466d65c9 ("mbuf: use 2 bytes for port and nb segments")
Cc: stable@dpdk.org
Signed-off-by: Ilya V. Matveychikov <matvejchikov@gmail.com>
Acked-by: Olivier Matz <olivier.matz@6wind.com>

net/sfc: avoid Tx queue setup failure if thresholds are set

Tx queue prefetch, host and writeback thresholds are used for
performance fine-tuning and not applicable to Solarflare NICs.
It is safe to just log warning and do not fail Tx queue setup
to be more friendly to DPDK applications which hardcode it.

Fixes: b1b7ad933b39 ("net/sfc: set up and release Tx queues")
Cc: stable@dpdk.org
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>

net/sfc: avoid Rx queue setup failure if thresholds are set

Rx queue prefetch, host and writeback thresholds are used for
performance fine-tuning and not applicable to Solarflare NICs.
It is safe to just log warning and do not fail Rx queue setup
to be more friendly to DPDK applications which hardcode it.

Fixes: ce35b05c635e ("net/sfc: implement Rx queue setup release operations")
Cc: stable@dpdk.org
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>

net/mlx5: fix build without soft counters

total_length is only visible when SOFT_COUNTERS are enabled

Fixes: 3f13f8c23a7c ("net/mlx5: support hardware TSO")
Cc: stable@dpdk.org
Signed-off-by: Thierry Herbelot <thierry.herbelot@6wind.com>
Acked-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>

net/mlx5: fix flow director deletion

The specification are wrongly computed causing the rule to not be match
correctly for deletion.

Fixes: 4c3e9bcdd52e ("net/mlx5: support flow director")
Signed-off-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
Acked-by: Yongseok Koh <yskoh@mellanox.com>

version: 17.11-rc4

Signed-off-by: Thomas Monjalon <thomas@monjalon.net>

doc: add tested platforms with Mellanox NICs

Signed-off-by: Raslan Darawsheh <rasland@mellanox.com>

doc: fix a typo in EAL guide

Fix an error in DPDK programmer's guide (EAL section):
it should be rte_thread_get_affinity() instead of
rte_pthread_get_affinity().

Signed-off-by: Rami Rosen <rami.rosen@intel.com>
Acked-by: John McNamara <john.mcnamara@intel.com>

doc: fix a typo in pipeline app guide

This patch fixes a trivial typo in ip pipeline app guide.

Signed-off-by: Rami Rosen <rami.rosen@intel.com>
Acked-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Acked-by: John McNamara <john.mcnamara@intel.com>

usertools: fix a typo in bind script

Signed-off-by: Pavel Shirshov <pavel.shirshov@gmail.com>

lib: fix typos

Signed-off-by: Pavel Shirshov <pavel.shirshov@gmail.com>

bus/pci: fix a typo in doxygen file description

Fixes: 764bf26873b9 ("add FreeBSD support")
Signed-off-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>

net/virtio: fix memory leak on failure

In function eth_virtio_dev_init(), dynamic memory stored
in "eth_dev->data->mac_addrs" variable and it is not freed
when function return,
this is a possible memory leak.

Fixes: 8ced1542f7a3 ("net/virtio: eth_dev->data->mac_addrs is not freed")
Cc: stable@dpdk.org
Signed-off-by: Pengzhen Liu <liupengzhen3@huawei.com>
Acked-by: Yuanhan Liu <yliu@fridaylinux.org>

ring: guarantee load/load order in enqueue and dequeue

We watched a rte panic of mbuf_autotest in our qualcomm arm64 server
(Amberwing).

Root cause:
In __rte_ring_move_cons_head()
...
        do {
                /* Restore n as it may change every loop */
                n = max;

                *old_head = r->cons.head;                //1st load
                const uint32_t prod_tail = r->prod.tail; //2nd load

In weak memory order architectures (powerpc,arm), the 2nd load might be
reodered before the 1st load, that makes *entries is bigger than we wanted.
This nasty reording messed enque/deque up.

cpu1(producer)          cpu2(consumer)          cpu3(consumer)
                        load r->prod.tail
in enqueue:
load r->cons.tail
load r->prod.head

store r->prod.tail

                                                load r->cons.head
                                                load r->prod.tail
                                                ...
                                                store r->cons.{head,tail}
                        load r->cons.head

Then, r->cons.head will be bigger than prod_tail, then make *entries very
big and the consumer will go forward incorrectly.

After this patch, the old cons.head will be recaculated after failure of
rte_atomic32_cmpset

There is no such issue on X86, because X86 is strong memory order model.
But rte_smp_rmb() doesn't have impact on runtime performance on X86, so
keep the same code without architectures specific concerns.

Fixes: 50d769054872 ("ring: add burst API")
Cc: stable@dpdk.org
Signed-off-by: Jia He <jia.he@hxt-semitech.com>
Signed-off-by: Jie Liu <jie2.liu@hxt-semitech.com>
Signed-off-by: Bing Zhao <bing.zhao@hxt-semitech.com>
Acked-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>
Acked-by: Jianbo Liu <jianbo.liu@arm.com>

maintainers: update PTP client example

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Acked-by: Daniel Mrzyglod <danielx.t.mrzyglod@intel.com>
Acked-by: John McNamara <john.mcnamara@intel.com>

maintainers: update job stats library

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Acked-by: Pawel Wodkowski <pawelx.wodkowski@intel.com>
Acked-by: John McNamara <john.mcnamara@intel.com>

maintainers: update for fm10k

Signed-off-by: Xiao Wang <xiao.w.wang@intel.com>
Acked-by: Helin Zhang <helin.zhang@intel.com>

maintainers: resign from mlx4 maintenance

I have been a little too busy these past months and could not follow all
the re-work of this PMD.

So the best thing for this PMD would be to move the mlx4 maintenance to
more involved people.

Cc: stable@dpdk.org
Signed-off-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>

maintainers: claim maintainership of VDEV bus

Signed-off-by: Jianfeng Tan <jianfeng.tan@intel.com>
Acked-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Zhiyong Yang <zhiyong.yang@intel.com>

maintainers: fill git trees for net and crypto API

The ethdev API (including rte_flow) is managed in the dpdk-next-net tree.
The crypto API is managed in the dpdk-next-crypto tree.

Signed-off-by: Thomas Monjalon <thomas@monjalon.net>

doc: mention 17.11 LTS in contributing guide

Signed-off-by: Luca Boccassi <bluca@debian.org>
Acked-by: John McNamara <john.mcnamara@intel.com>
Acked-by: Thomas Monjalon <thomas@monjalon.net>

doc: update API/ABI policy

Following agreement at the DPDK Technical Board meeting of 2017-10-13 [1],
update the documentation with the ABI/API policy changes.

[1] http://dpdk.org/ml/archives/dev/2017-October/079961.html

Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: John McNamara <john.mcnamara@intel.com>
Acked-by: Yuanhan Liu <yliu@fridaylinux.org>
Acked-by: Kevin Traynor <ktraynor@redhat.com>
Acked-by: Thomas Monjalon <thomas@monjalon.net>

doc: update deprecation of ethdev offload API

Update deprecation notice for the new ethdev offloads API.
Deprecation of the old offloads API is set to 18.05.

Signed-off-by: Shahaf Shuler <shahafs@mellanox.com>
Acked-by: Andrew Rybchenko <arybchenko@solarflare.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
Acked-by: Thomas Monjalon <thomas@monjalon.net>

doc: postpone devargs clean-up

These changes were planned for 17.11 but were proposed too late.
Postpone those to v18.02 instead.

Signed-off-by: Gaetan Rivet <gaetan.rivet@6wind.com>

doc: update release notes for 17.11

Fix grammar, spelling and formatting of DPDK 17.11 release notes.

Signed-off-by: John McNamara <john.mcnamara@intel.com>

doc: add tested Intel platforms with Intel NICs

Add tested Intel platforms with Intel NICs to the release note.

Signed-off-by: Yulong Pei <yulong.pei@intel.com>

doc: fix mempool option in octeontx guide

Fixes: f820b5896631 ("doc: add octeontx ethdev driver documentation")
Signed-off-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Acked-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>
Acked-by: John McNamara <john.mcnamara@intel.com>

doc: add octeontx mempool guide

This commit adds a section to the docs listing the mempool
device PMDs available.

It then adds the octeontx fpavf mempool PMD to the listed mempool
devices.

Signed-off-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Acked-by: John McNamara <john.mcnamara@intel.com>

doc: add octeontx platform guide

This commit adds a section to the docs listing the platform
guide for the PMDs.

It then adds the octeontx platform guide to the listed platform
devices.

Patch also removes platform specific duplicate setup information from
eventdev/octeontx.rst, nics/octeontx.rst and update to
plaform/octeontx.rst.

Signed-off-by: Santosh Shukla <santosh.shukla@caviumnetworks.com>
Acked-by: John McNamara <john.mcnamara@intel.com>

doc: update mlx4 documentation

This updates mlx4 documentation and DPDK release notes
to reflect the PMD support for rdma-core from linux-rdma.

- PMD is now freed from Mellanox OFED and now only depends on the
  public rdma-core package (v15 and above) instead.
  (see https://github.com/linux-rdma/rdma-core/releases)
  This PMD should run under Linux v4.14 and above.
- In case any of the above requirements can't be satisfied,
  Mellanox OFED v4.2 and above also provide an updated rdma-core
  as well back-ported kernel modules for most Linux distributions
  and previous Linux versions.
  (see http://www.mellanox.com/page/products_dyn?product_family=26).

Signed-off-by: Moti Haimovsky <motih@mellanox.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>

doc: remove mlx4 Rx inline support

This commit removes the stale MLX4_INLINE_RECV_SIZE environment
variable from the documentation.

Fixes: 056eaf2e6d55 ("net/mlx4: drop inline receive support")
Signed-off-by: Moti Haimovsky <motih@mellanox.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>

doc: remove mlx4 Tx inline compilation option

This patch removes the stale CONFIG_RTE_LIBRTE_MLX4_MAX_INLINE compilation
option from the documentation.

Fixes: 586db080585b ("net/mlx4: remove Tx inline compilation option")
Signed-off-by: Moti Haimovsky <motih@mellanox.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>

doc: add inline crypto feature

Signed-off-by: Radu Nicolau <radu.nicolau@intel.com>

doc: move fast mbuf free feature in net guide

The feature was added at the end of the table.
And the description was between the anchor _nic_features_timesync
and its title.
It is moved near related features with a new anchor.

It is also renamed from "mbuf fast free" to "fast mbuf free".

Fixes: d6f90afd3070 ("ethdev: add mbuf fast free Tx offload")
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Ferruh Yigit <ferruh.yigit@intel.com>

pdump: fix possible mbuf leak on failure

If pdump_pktmbuf_copy_data() fails it's possible to have segment leak
as rte_pktmbuf_free() only handles m_dup chain but not the seg just
allocated and yet not chained.

Fixes: 278f945402c5 ("pdump: add new library for packet capture")
Signed-off-by: Ilya V. Matveychikov <matvejchikov@gmail.com>

examples/ipv4_multicast: fix segments number type

Fixes: 97cb466d65c9 ("mbuf: use 2 bytes for port and nb segments")
Signed-off-by: Ilya V. Matveychikov <matvejchikov@gmail.com>

test/memory: fix autotest parsing

This patch fixes three problems in memory autotest:
 - the regex for IOVA
 - missing validation of second and following output lines
 - propagation of error to consecutive tests
 - conversion base for mem size (hex indtead of dec)
First fix is for changes introduced with IOVA, the regex was not
updated which lead to unit test failure.  Patch now also uses loop for
line outputs processing to verify more than just one line. By this we
also satisfy the pexpect() and scan the "Test OK"/"Test Failed" so in
case of error all output lines are consumed and does not break the
consecutive test (error does not propagate).

Fixes: 7ba49d39f14c ("mem: rename segment address from physical to IOVA")
Fixes: b4ce15aa2bc7 ("app/test: fix memory autotest")
Signed-off-by: Radoslaw Biernacki <radoslaw.biernacki@linaro.org>

test/memzone: fix memory leak

This patch fixes the memory leaks in memzone_autotest. Those memory leaks
lead to failures in tests from the same testing group due to out of memory
problems.  With introduction of rte_memzone_free() it is now possible to
free the memzone.  Fix uses this API call to make a clean after each test
case.

Fixes: ff909fe21f0a ("mem: introduce memzone freeing")
Cc: stable@dpdk.org
Signed-off-by: Radoslaw Biernacki <radoslaw.biernacki@linaro.org>

app/testpmd: remove port status check from TM command

Currently, testpmd CLI doesn't permit to add leaf and non-leaf node when
port is started. It doesn't work in case of i40e device as DCB
configuration is deleted when port is stopped. Therefore, removes the
port status check before invoking leaf and nonleaf node API in the cli.
If needed, device can add port status check at the driver layer.

Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>
Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>

eal: fix an include guard comment

Fixes: af75078fece3 ("first public release")
Signed-off-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>

net/virtio: fix Rx interrupt with VFIO

When running l3fwd-power to test virtio rxq interrupt using vfio
pci noiommu mode, startup fails. In the function virtio_read_caps,
the code if (flags & PCI_MSIX_ENABLE) intends to double check
if vfio msix is enabled or not. However, it is not enable at that
time. So use_msix is assigned to "0", not "1", which causes the
failure of configuring rxq intr in l3fwd-power.
This patch adds the function "vtpci_msix_detect" to detect the status
of msix when interrupt changes happen.
In the meanwhile, virtio_intr_enable/disable are introduced to wrap
rte_intr_enable/disable to enhance the ability to detect msix.
use_msix can indicate three different msix status by:
VIRTIO_MSIX_NONE (0)
VIRTIO_MSIX_DISABLED (1)
VIRTIO_MSIX_ENABLED (2)

Fixes: cb482cb3a305 ("net/virtio: fix MAC address read")
Cc: stable@dpdk.org
Signed-off-by: Zhiyong Yang <zhiyong.yang@intel.com>
Acked-by: Jianfeng Tan <jianfeng.tan@intel.com>
Acked-by: Maxime Coquelin <maxime.coquelin@redhat.com>

net/mlx5: fix link speed bitmasks

The constant ETHTOOL_LINK_MODE_1000baseT_Full_BIT and the others like
that in mlx5_link_update_unlocked_gs must be bit masks but unfortunately
they are bit numbers. This commit fixes the issue.

Fixes: 188408719888 ("net/mlx5: fix support for newer link speeds")
Cc: stable@dpdk.org
Signed-off-by: Edward Makarov <makarov@kraftway.ru>
Acked-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>

net/mlx4: fix missing stamp during Tx completion

After processing completed packets, the owner bit of each TXBB comprised
in its WQEs must be invalidated. The loop stops short of processing the
last WQE.

Fixes: c3c977bbecbd ("net/mlx4: add Tx bypassing Verbs")
Signed-off-by: Matan Azrad <matan@mellanox.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>

net/bnxt: fix link handling and configuration

Remove a case where we were sending a deprecated field to the FW.
There is no need to send auto_link_speed to the FW.
Also set the auto_mode correctly depending on the setting requested.

Fixes: 7bc8e9a227cc ("net/bnxt: support async link notification")
Cc: stable@dpdk.org
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/bnxt: fix duplicate creation of ntuple filter

Prevent the creation of duplicate 5tuple filters.

Fixes: b7435d660a8c ("net/bnxt: add ntuple filtering support")
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>

net/nfp: fix possible memory leak

Memory allocated was not being released in any exit path.

Coverity issue: 195030
Fixes: 48e2255f1b63 ("net/nfp: add NSP support for HW link configuration")
Signed-off-by: Alejandro Lucero <alejandro.lucero@netronome.com>

net/nfp: fix null pointer check

First, the received pointer was not checked before. Then the pointer
from malloc was not the one used in the existing check.

Coverity issue: 195027
Fixes: ad60bca34899 ("net/nfp: read PF port MAC addr using NSP")
Signed-off-by: Alejandro Lucero <alejandro.lucero@netronome.com>

net/nfp: fix possible bad shif operation

We do not know how big can the BAR be, but we know anything less
than 1MB is an error. This BAR needs to be big enough for accessing
most of NFP internals.

Coverity issue: 195024
Fixes: d12206e00590 ("net/nfp: add NSP user space interface")
Signed-off-by: Alejandro Lucero <alejandro.lucero@netronome.com>

net/nfp: fix checking function return value

The fstat function could return a value that indicates an error condition.
If this is not checked, the error condition may not be handled correctly.

Coverity issue: 195019
Fixes: f37d8a4b67b2 ("net/nfp: add NSP FW upload command")
Signed-off-by: Alejandro Lucero <alejandro.lucero@netronome.com>

net/nfp: fix resource leak

File descriptor is not released in any potential exit path
inside the function.

Coverity issue: 195018
Fixes: f37d8a4b67b2 ("net/nfp: add NSP FW upload command")
Signed-off-by: Alejandro Lucero <alejandro.lucero@netronome.com>

net/nfp: fix memory allocation

If the function actually returns a null value, a null pointer
dereference will occur.

Coverity issue: 195013
Fixes: dd63df2bfff3 ("net/nfp: add NSP symbol resolution command")
Signed-off-by: Alejandro Lucero <alejandro.lucero@netronome.com>

net/nfp: fix stats struct initial value

Not all struct fields will be written and random data could
confuse readers.

Coverity issue: 140755
Fixes: 92aa491b881e ("nfp: add statistics")
Cc: stable@dpdk.org
Signed-off-by: Alejandro Lucero <alejandro.lucero@netronome.com>

drivers/net: fix number of segment storage type

Fixes: 97cb466d65c9 ("mbuf: use 2 bytes for port and nb segments")
Cc: stable@dpdk.org
Signed-off-by: Ilya V. Matveychikov <matvejchikov@gmail.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>

net/mlx4: fix rxq interrupt memory corruption

intr_vec allocation size was wrong causing a memory corruption.

Fixes: 0a2ae703192c ("net/mlx4: fix Rx interrupts management")
Cc: stable@dpdk.org
Signed-off-by: Shahaf Shuler <shahafs@mellanox.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>

net/mlx5: fix rxq interrupt memory corruption

intr_vec allocation size was wrong causing a memory corruption.

Fixes: e1016cb73383 ("net/mlx5: fix Rx interrupts management")
Cc: stable@dpdk.org
Signed-off-by: Shahaf Shuler <shahafs@mellanox.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>

net/mlx5: fix flow director rules comparison

When deleting/replacing a flow director rule the first rule of the list
was always the one replaced instead of the corresponding one.

Fixes: 4c3e9bcdd52e ("net/mlx5: support flow director")
Signed-off-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>

net/qede: fix default config option

Restore the default configuration as in previous releases and
add a debug msg.

Fixes: f07aa795c92a ("net/qede: disable per-VF Tx switching feature")
Cc: stable@dpdk.org
Signed-off-by: Harish Patil <harish.patil@cavium.com>
Signed-off-by: Rasesh Mody <rasesh.mody@cavium.com>

doc: update qede PMD NIC guide

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

net/mlx4: fix Rx packet type offloads

This patch improves Rx packet type offload report in case the device is
a virtual function device.
In these devices we observed that the L2 tunnel flag is set also for
non-tunneled packets, this leads to a complete misinterpretation of the
packet type being received.
This issue occurs since the tunnel_mode is not set to 0x7 by the driver
for virtual devices and therefore the value in the L2 tunnel flag is
meaningless and should be ignored.

Fixes: aee4a03fee4f ("net/mlx4: enhance Rx packet type offloads")
Signed-off-by: Moti Haimovsky <motih@mellanox.com>
Acked-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>

net/i40e: fix VF cannot forward packets issue

When Rx interrupt is not enabled, there is no need to check if interrupt
allow others. It will cause VF cannot forwarding packets issue. This patch
fixes this issue.

Fixes: 96a9fd03c25f ("net/i40e: fix Rx queue interrupt mapping in VF")
Cc: stable@dpdk.org
Signed-off-by: Xiaoyun Li <xiaoyun.li@intel.com>
Acked-by: Beilei Xing <beilei.xing@intel.com>

security: fix anonymous unions in structure definitions

Fixes: c261d1431bd8 ("security: introduce security API and framework")
Signed-off-by: Akhil Goyal <akhil.goyal@nxp.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

examples/ipsec-secgw: fix build with gcc 4.5.1

On SUSE11 with gcc 4.5.1 the following build error occurred,
most likely because of improper handling of annonymous unions:

examples/ipsec-secgw/ipsec.c:87:4 error:
unknown field ‘ipsec’ specified in initializer

Fixes: ec17993a145a ("examples/ipsec-secgw: support security offload")
Signed-off-by: Radu Nicolau <radu.nicolau@intel.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

net/liquidio: support device reset in driver

Reset device during init and close if bound to igb_uio.

Fixes: 369db3ae8e91 ("igb_uio: remove device reset in release")
Cc: stable@dpdk.org
Signed-off-by: Shijith Thotton <shijith.thotton@caviumnetworks.com>

net/softnic: fix build with gcc 4.5.1

Fix the build error due to improper handling of unions
on SUSE11 (gcc 4.5.1).

drivers/net/softnic/rte_eth_softnic_tm.c:588:3
error: unknown field 'nonleaf' specified in initializer

Fixes: 299a89de916a ("net/softnic: add TM capabilities ops")
Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>
Acked-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>

version: 17.11-rc3

Signed-off-by: Thomas Monjalon <thomas@monjalon.net>

doc: adjust column width of PMD overview tables

The net PMD overview table is big and required to be shrunk.
The crypto PMD tables had variable column sizes: the latest
columns were smaller.

The minimum width is set to 0.9em without any forced padding.
The maximum width is set to 2em.
The width of the first column set by sphinx is overridden to 0
for full elasticity.

Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: John McNamara <john.mcnamara@intel.com>

doc: move dom0 removal notice in removed items

Since the release notes has a new section for removed items,
the dom0 removal notice can be moved there.

Signed-off-by: Thomas Monjalon <thomas@monjalon.net>

doc: add DPAA2 event adapter to release notes

Signed-off-by: Shreyansh Jain <shreyansh.jain@nxp.com>
Acked by: Nipun Gupta <nipun.gupta@nxp.com>
Acked-by: John McNamara <john.mcnamara@intel.com>

doc: add security library to release notes

Removed the deprication notice for ABI breakage and updated
release notes for rte_security.

Signed-off-by: Akhil Goyal <akhil.goyal@nxp.com>

doc: update IPsec multi-buffer library version

A new library of the IPsec Multi-buffer library has been
released (0.47), which is used by the AESNI GCM PMD.

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

doc: fix a typo in EAL guide

This patch fixes a trivial typo in DPDK programmer's guide:
it should be rte_cpu_get_features() instead of rte_cpu_get_feature().

Signed-off-by: Rami Rosen <rami.rosen@intel.com>
Acked-by: John McNamara <john.mcnamara@intel.com>