net/softnic: add run API

Implements softnic API function to run pipeline objects.

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

net/softnic: add thread

Add thread data structure and init function to run softnic pipelines
objects.

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

net/softnic: add pipeline object

Add pipeline object implementation to the softnic.

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

net/softnic: add table action profile

Add pipeline's table action profile implementation to the softnic.

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

net/softnic: add port action profile

Add pipeline's port action profile implementation to the softnic.

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

net/softnic: add traffic manager object

Add traffic manager(tmgr) object to the softnic.

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

net/softnic: add tap object

Add tap object implementation to the softnic.

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

net/softnic: add mempool object

Add mempool object implementation to the softnic.

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

net/softnic: add link object

Add link object implementation to the softnic.

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

net/softnic: add software queue object

Add swq object implementation to the softnic.

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

net/softnic: restructure

Rework the softnic implementation to have flexiblity in enabling
more features to its receive and transmit data path.

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

examples/ip_pipeline: add link command

Add the functionality to track links in the application. This enables the
user to print the name, mac address and statistics for each link
in the application.

Signed-off-by: Kevin Laatz <kevin.laatz@intel.com>
Acked-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>

raw/dpaa2_qdma: fix IOVA as VA flag

Fixes: b1ee472fed58 ("raw/dpaa2_qdma: introduce the DPAA2 QDMA driver")
Cc: stable@dpdk.org
Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>

bus/dpaa: support scatter/gather config

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>

net/dpaa: move the push queue set to global init

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com>

drivers: support function name in NXP logs

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com>

bus/fslmc: cleanup unnecessary global variables

Signed-off-by: Pavan Nikhilesh <pbhagavatula@caviumnetworks.com>
Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>

bus/dpaa: cleanup unnecessary global variables

Signed-off-by: Pavan Nikhilesh <pbhagavatula@caviumnetworks.com>
Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>

net/dpaa: remove experimental tag from PMD API

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com>

net/dpaa: support default queue mode

In case DPAA FMAN configuration tool (FMC) is not available,
the system can still work with default queue(1 queue per port).

This patch also fixes some logs related to FQ ids, which were
idetified while testing this support.

Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>

bus/dpaa: make vdqcr configurable

This patch add support for configurable vdqcr exact flag.
This boost the performance, however this can give
side effects for some extra packet fetch. Which has been
taken care in the patch as well.

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

bus/dpaa: add API to get MAC address

Signed-off-by: Akhil Goyal <akhil.goyal@nxp.com>
Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com>

bus/dpaa: optimize fq callback routine

Avoid array of fq as packets are dq only from a single q.

Signed-off-by: Sunil Kumar Kori <sunil.kori@nxp.com>
Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com>

net/dpaa2: fix prefetch Rx to honor number of packets

This patch fixes prefetch rx routine to
set the next prefetch request to the size of nb_pkts.
It assumes that next request would ideally be
of same size.

Fixes: 4bc5ab88dbd6 ("net/dpaa2: fix Tx only mode")
Cc: stable@dpdk.org
Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com>

net/dpaa: fix queue error handling and logs

Fixes: 5e7455931442 ("net/dpaa: support Rx queue configurations with eventdev")
Cc: stable@dpdk.org
Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com>

bus/dpaa: fix buffer offset setting in FMAN

The buffer offset was incorrectly being set at 64,
thus not honoring the packet headroom.

Fixes: 6d6b4f49a155 ("bus/dpaa: add FMAN hardware operations")
Cc: stable@dpdk.org
Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com>

bus/dpaa: fix SVR id fetch location

Otherwise the SVR may not be available for dpaa init.

Fixes: 3b59b73dea08 ("bus/dpaa: update platform SoC value register routines")
Cc: stable@dpdk.org
Signed-off-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com>

bus/dpaa: fix phandle support for Linux 4.16

Fixes: 2183c6f69d7e ("bus/dpaa: add OF parser for device scanning")
Cc: stable@dpdk.org
Signed-off-by: Alok Makhariya <alok.makhariya@nxp.com>
Acked-by: Shreyansh Jain <shreyansh.jain@nxp.com>

mk: change TLS model for DPAA machine

Random corruptions observed on platfoms with using
the dpdk library in shared mode with VPP software (plugin).

using traditional TLS scheme resolved the issue.

Tested with VPP with DPDK as a plugin.

Signed-off-by: Sachin Saxena <sachin.saxena@nxp.com>

hash: validate hash bucket entries while compiling

Validate RTE_HASH_BUCKET_ENTRIES during compilation instead of
run time.

Signed-off-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
Reviewed-by: Gavin Hu <gavin.hu@arm.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

net/kni: fix check for meson build

The configuration value indicating that KNI was build was incorrect,
causing the driver to never be built.

Fixes: 3479586fe636 ("net/kni: add to meson build")

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

net/avp: fix 32-bit meson builds

When compiling with meson, extra warnings are enabled about casting from
integers to different size pointers. This triggers an error in AVP as the
addition of the offset to the pointer address causes the result to be a
64-bit integer which doesn't fit a 32-bit pointer. The fix here is to
explicitly indicate that the offset is of type "uintptr_t" which prevents
any promotion which would cause errors.

Fixes: c0ad584222b5 ("net/avp: add device initialization")

Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: Allain Legacy <allain.legacy@windriver.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>

examples/kni: fix dependency check for meson build

Rather than hard-coding the example app to be built only when a set of
conditions are met, we can simplify things by having the app built when
KNI library itself is available. That saves us duplicating the same set
of restrictions on both library and example app.

Fixes: 89f0711f9ddf ("examples: build some samples with meson")

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

bus/dpaa2: fix default IOVA setting for meson builds

By default, the DPAA2_USE_PHYS_IOVA setting with the make build-system
is "y", which is overridden to "n" in the config file specifically for
the arm64-dpaa2-linuxapp-gcc builds. The lack of this setting in meson
builds causes issues for 32-bit build, as the alternative code paths
have compiler warnings e.g. for i686 builds. Therefore we should align
the meson and make settings, setting the value to "true" by default and
overriding it to "false" for dpaa2-specific builds.

Fixes: 6ec78c2463ac ("build: add meson support for dpaaX platforms")

Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>

build: disable pointer to int warnings for 32-bit

Converting a 32-bit pointer to a 64-bit integer is generally safe, but
generates a lot of warnings when compiling 32-bit code with meson. The
warnings are not flagged when using make, so just disable them for
32-bit meson builds.

Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>

net/sfc: disable for 32-bit builds

The sfc driver is not supported on 32-bit so disable in meson in
those cases.

Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: Andrew Rybchenko <arybchenko@solarflare.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>

bpf: fix 32-bit build support with meson

The JIT is only supported on x86_64 so disable for 32-bit builds.

Fixes: cc752e43e079 ("bpf: add JIT compilation for x86_64 ISA")

Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>

kni: disable for 32-bit meson builds

The kni library is not supported on 32-bit so disable it for those
builds using meson.

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

examples: make Linux environment check consistent

Some Makefiles are using CONFIG_RTE_EXEC_ENV and others
are using CONFIG_RTE_EXEC_ENV_LINUXAPP.
Use the latter one for consistency.
We could remove CONFIG_RTE_EXEC_ENV later if considered useless.

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

test/bpf: add test for external function call

Signed-off-by: Konstantin Ananyev <konstantin.ananyev@intel.com>

bpf: add extra validation for input BPF program

Add checks for:
 - use/return uninitialized registers and/or stack data
 - possible memory access boundaries violation
 - invalid arguments for the function

Signed-off-by: Konstantin Ananyev <konstantin.ananyev@intel.com>

bpf: add extra information for external symbol definitions

Extend struct rte_bpf_xsym with new fields to provide information about:
 - for variables - type and size
 - for functions - number of arguments and type/size of each argument
   and return value

Such information would allow validate code to perform
more extensive checking on input BPF program and catch
misbehaving BPF code.

That change would cause ABI/API breakage for librte_bpf.

Signed-off-by: Konstantin Ananyev <konstantin.ananyev@intel.com>

eventdev: add callback for Rx adapter SW transfers

Add ability for application to register a callback function
for SW transfers, the callback can decide which packets can
be enqueued to the event device.

Signed-off-by: Nikhil Rao <nikhil.rao@intel.com>
Acked-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>

test/event: add Rx adapter tests for interrupt driven queues

Add test for queue add and delete, the add/delete calls
also switch queues between poll and interrupt mode.

Signed-off-by: Nikhil Rao <nikhil.rao@intel.com>

eventdev: add interrupt driven queues to Rx adapter

Add support for interrupt driven queues when eth device is
configured for rxq interrupts and servicing weight for the
queue is configured to be zero.

A interrupt driven packet received counter has been added to
rte_event_eth_rx_adapter_stats.

Signed-off-by: Nikhil Rao <nikhil.rao@intel.com>

eventdev: move Rx adapter to separate function

Create a separate function that handles eth receive and
enqueue to event buffer. This function will also be called for
interrupt driven receive queues.

Signed-off-by: Nikhil Rao <nikhil.rao@intel.com>
Acked-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>

eventdev: improve error handling for Rx adapter queue add/del

The new WRR sequence applicable after queue add/del is set
up after setting the new queue state, so a memory allocation
failure will leave behind an incorrect state.

This change separates the memory sizing + allocation for the
Rx poll and WRR array from calculation of the WRR sequence.
If there is a memory allocation failure, existing Rx queue
configuration remains unchanged.

Signed-off-by: Nikhil Rao <nikhil.rao@intel.com>

eventdev: standardize Rx adapter internal function names

Add a common prefix to function names and rename
few to better match functionality

Signed-off-by: Nikhil Rao <nikhil.rao@intel.com>
Acked-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>

event/sw: support device stop flush callback

This commit also adds a flush callback test to the sw eventdev's selftest
suite.

Signed-off-by: Gage Eads <gage.eads@intel.com>
Acked-by: Harry van Haaren <harry.van.haaren@intel.com>

service: add mechanism for quiescing

Existing service functions allow us to stop a service, but doing so doesn't
guarantee that the service has finished running on a service core. This
commit introduces rte_service_may_be_active(), which returns whether the
service may be executing on one or more lcores currently, or definitely is
not.

The service core layer supports this function by setting a flag when
a service core is going to execute a service, and unsetting the flag when
the core is no longer able to run the service (its runstate becomes stopped
or the lcore is no longer mapped).

With this new function, applications can set a service's runstate to
stopped, then poll rte_service_may_be_active() until it returns false. At
that point, the service is quiesced.

Signed-off-by: Gage Eads <gage.eads@intel.com>
Acked-by: Harry van Haaren <harry.van.haaren@intel.com>

event/octeontx: fix flush callback

When event queues are being flushed the getwork operation used to extract
events should be a grouped getwork operation to the specific event queue.

Fixes: 8384f0e039ea ("event/octeontx: support device stop flush callback")
Cc: stable@dpdk.org
Signed-off-by: Pavan Nikhilesh <pbhagavatula@caviumnetworks.com>
Acked-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>

eventdev: include PCI devices when getting id from name

Currently, rte_event_dev_get_dev_id works only for virtual devices(vdev)
update the function to include devices probed through pci.

Signed-off-by: Pavan Nikhilesh <pbhagavatula@caviumnetworks.com>
Acked-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>

remove useless constructor headers

A constructor is usually declared with RTE_INIT* macros.
As it is a static function, no need to declare before its definition.
The macro is used directly in the function definition.

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

doc: add UDP/IPv4 GSO in guides

This patch updates the programmer guide and testpmd user guide for
UDP/IPv4 GSO.

Signed-off-by: Jiayu Hu <jiayu.hu@intel.com>
Acked-by: Xiao Wang <xiao.w.wang@intel.com>

app/testpmd: enable UDP GSO in csum engine

This patch enables GSO for UDP/IPv4 packets. Oversized UDP/IPv4
packets transmitted over a GSO-enabled port will undergo segmentation.

Signed-off-by: Jiayu Hu <jiayu.hu@intel.com>
Acked-by: Xiao Wang <xiao.w.wang@intel.com>
Acked-by: Bernard Iremonger <bernard.iremonger@intel.com>
Tested-by: Yuwei Zhang <yuwei1.zhang@intel.com>

gso: support UDP/IPv4 fragmentation

This patch adds GSO support for UDP/IPv4 packets. Supported packets
may include a single VLAN tag. UDP/IPv4 GSO doesn't check if input
packets have correct checksums, and doesn't update checksums for
output packets (the responsibility for this lies with the application).
Additionally, UDP/IPv4 GSO doesn't process IP fragmented packets.

UDP/IPv4 GSO uses two chained MBUFs, one direct MBUF and one indrect
MBUF, to organize an output packet. The direct MBUF stores the packet
header, while the indirect mbuf simply points to a location within the
original packet's payload. Consequently, use of UDP GSO requires
multi-segment MBUF support in the TX functions of the NIC driver.

If a packet is GSO'd, UDP/IPv4 GSO reduces its MBUF refcnt by 1. As a
result, when all of its GSOed segments are freed, the packet is freed
automatically.

Signed-off-by: Jiayu Hu <jiayu.hu@intel.com>
Acked-by: Xiao Wang <xiao.w.wang@intel.com>

service: add attribute API

Add APIs that allow an application to query and reset the attributes of
a service lcore.  Add one such new attribute, "loops", which is a
counter that tracks the number of times the service core has looped in
the service runner function.  This is useful to applications that desire
a "liveness" check to make sure a service core is not stuck.

Signed-off-by: Erik Gabriel Carrillo <erik.g.carrillo@intel.com>
Acked-by: Harry van Haaren <harry.van.haaren@intel.com>

doc: add default that all fixes should be backported

Set the starting point that all commits on master branch
with Fixes tag should be backported to relevant stable/LTS
branches, and explain that the submitter may indicate it is
not suitable for backport.

Of course there will be exceptions that will crop up from time
to time that need discussion, so also add a sentence for that.

This is to ensure that there is consistency between what is
backported to stable/LTS branches, remove some subjectivity
as to what constitutes "a fix" and avoid possible conflicts
for future backports.

Signed-off-by: Kevin Traynor <ktraynor@redhat.com>
Acked-by: Luca Boccassi <bluca@debian.org>
Acked-by: Aaron Conole <aconole@redhat.com>
Acked-by: Thomas Monjalon <thomas@monjalon.net>

devtools: expand meson cross compiling test coverage

The default test script covers only default host cc compiler, either gcc or
clang, the fix is to increase the coverage by adding one more to cover
clang and the others for gcc.

Fixes: a55277a788 ("devtools: add test script for meson builds")
Cc: stable@dpdk.org
Signed-off-by: Gavin Hu <gavin.hu@arm.com>
Reviewed-by: Phil Yang <phil.yang@arm.com>
Reviewed-by: Song Zhu <song.zhu@arm.com>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>

build: fix for host clang and cross gcc

The following error hits if host cc compiler is clang(default one in most
linux distributions) and the cross compiler is gcc.

The root cause is: the hybride compilers add the warning options to the
meson project as project arguments, which apply for both host compiling and
cross compiling. But some options such as '-Wno-format-truncation' are not
supported nor recognized by clang, so they have to be removed from the
project arguments for the host compiler to run smoothily and added back as
cflags for the cross compiler to compile for cross source files.

The fix is remove unrecognized warning options from the meson project
arguments shared by gcc and clang, as add them specifically for gcc or
clang as cflags.

[265/893] Compiling C object
'buildtools/pmdinfogen/pmdinfogen@exe/pmdinfogen.c.o'.  warning: unknown
warning option '-Wno-format-truncation' [-Wunknown-warning-option]

Fixes: a55277a788 ("devtools: add test script for meson builds")
Cc: stable@dpdk.org
Signed-off-by: Gavin Hu <gavin.hu@arm.com>
Reviewed-by: Phil Yang <phil.yang@arm.com>
Reviewed-by: Song Zhu <song.zhu@arm.com>
Reviewed-by: Steve Capper <steve.capper@arm.com>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>

devtools: fix ninja command in build test

On some linux distributions, eg: CentOS, the ninja executable has a
different name: ninja-build, this patch is to check and adapt to it
accordingly.

./devtools/test-meson-builds.sh: line 24: ninja: command not found

Fixes: a55277a788 ("devtools: add test script for meson builds")
Cc: stable@dpdk.org
Signed-off-by: Gavin Hu <gavin.hu@arm.com>
Reviewed-by: Phil Yang <phil.yang@arm.com>
Reviewed-by: Song Zhu <song.zhu@arm.com>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>

build: fix meson warning

This is to fix the unnecessary warning output, it is not consistent with
the configurations of other platforms.

WARNING: Cross file does not specify strip binary, result will not be
stripped.

Fixes: e53a5299d2 ("build: support vendor specific ARM cross builds")
Cc: stable@dpdk.org
Signed-off-by: Gavin Hu <gavin.hu@arm.com>
Reviewed-by: Phil Yang <phil.yang@arm.com>
Reviewed-by: Song Zhu <song.zhu@arm.com>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>
Tested-by: Ali Alnubani <alialnu@mellanox.com>

doc: add cross compiling guide

This is the guide for cross compiling ARM64 DPDK from X86 hosts.

Signed-off-by: Gavin Hu <gavin.hu@arm.com>
Reviewed-by: Steve Capper <steve.capper@arm.com>
Reviewed-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
Acked-by: Marko Kovacevic <marko.kovacevic@intel.com>

mk: fix cross build

The "-Wimplicit-fallthrough=2" option was introduced into gcc 7.0, it was
enabled when the cross compiler gcc is greater than 7.0, but for the host
side buildtools/pmdinfogen, if the native gcc is older than 7.0, or the
host cc compiler is clang, it should not be enabled.

The fix is to differentiate the host gcc Werror options from the cross gcc.

gcc -Wp,-MD,./.pmdinfogen.o.d.tmp  -W -Wall -Wstrict-prototypes
-Wmissing-prototypes -Wmissing-declarations -Wold-style-definition
-Wpointer-arith -Wcast-align -Wnested-externs -Wcast-qual
-Wformat-nonliteral -Wformat-security -Wundef -Wwrite-strings -Wdeprecated
-Werror -Wimplicit-fallthrough=2 -Dbbb -Wno-format-truncation -g
-I/dpdk/build/include    -o pmdinfogen.o -c
~/dpdk/buildtools/pmdinfogen/pmdinfogen.c gcc: error:
unrecognized command line option ‘-Wimplicit-fallthrough=2’

Fixes: ced3e6f8 ("mk: adjust gcc flags for new gcc 7 warnings")
Cc: stable@dpdk.org
Signed-off-by: Gavin Hu <gavin.hu@arm.com>
Reviewed-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
Reviewed-by: Steve Capper <steve.capper@arm.com>
Tested-by: Ali Alnubani <alialnu@mellanox.com>
Acked-by: Thomas Monjalon <thomas@monjalon.net>

examples/multi_process: build l2fwd_fork app

l2fwd_fork is not complied by default, this will make it compile

Fixes: 95e8005a56e8 ("examples/l2fwd_fork: new app")

Signed-off-by: Emma Kenny <emma.kenny@intel.com>
Acked-by: Ferruh Yigit <ferruh.yigit@intel.com>

examples/multi_process: fix build

Fix bug with undeclared variable name and
calling a variable that is not member of struct.

CC main.o
l2fwd_fork/main.c:  In function ‘main’: l2fwd_fork/main.c:1043:33:
 error: ‘dev_info’ undeclared (first use in this function)
 rte_eth_dev_info_get(portid, &dev_info);

l2fwd_fork/main.c:1043:33: note: each undeclared identifier is
reported only once for each function it appears in
l2fwd_fork/main.c:1077:11: error: ‘struct rte_eth_txconf’
has no member named ‘tx_offloads’
   txq_conf.tx_offloads = local_port_conf.txmode.offloads;

Fixes: f8c02ca878af ("examples/multi_process: convert to new ethdev offloads API")
Cc: stable@dpdk.org
Signed-off-by: Emma Kenny <emma.kenny@intel.com>
Acked-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Shahaf Shuler <shahafs@mellanox.com>

maintainers: update for Mellanox PMDs

Shahaf and Matan volunteered to replace Nélio and myself as maintainers for
mlx4 and mlx5 PMDs. Cheers!

Cc: stable@dpdk.org
Signed-off-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
Signed-off-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com>
Acked-by: Shahaf Shuler <shahafs@mellanox.com>
Acked-by: Matan Azrad <matan@mellanox.com>
Acked-by: Yongseok Koh <yskoh@mellanox.com>

maintainers: update for mempool

Given its very good contributions to this library, add Andrew as
official maintainer for librte_mempool.

Signed-off-by: Olivier Matz <olivier.matz@6wind.com>
Acked-by: Andrew Rybchenko <arybchenko@solarflare.com>

igb_uio: fix build with lock down checks

Caught on ubuntu-16.04 with hwe kernel for aarch64:

$ uname -a
Linux ubuntu1604arm64es 4.13.0-43-generic #48~16.04.1-Ubuntu SMP Thu May
17 13:08:01 UTC 2018 aarch64 aarch64 aarch64 GNU/Linux

== Build kernel/linux/igb_uio
  CC [M] .../kernel/linux/igb_uio/igb_uio.o
In file included from .../kernel/linux/igb_uio/igb_uio.c:20:0:
.../igb_uio/compat.h: In function ‘igbuio_kernel_is_locked_down’:
.../igb_uio/compat.h:146:7:
error: "CONFIG_EFI_SECURE_BOOT_LOCK_DOWN" is not defined [-Werror=undef]
 #elif CONFIG_EFI_SECURE_BOOT_LOCK_DOWN
       ^
cc1: all warnings being treated as errors

Fixes: d67014c3d38b ("igb_uio: fail and log if kernel lock down is enabled")

Signed-off-by: David Marchand <david.marchand@6wind.com>
Acked-by: Ferruh Yigit <ferruh.yigit@intel.com>

test/crypto: add DH and DSA tests

Test application include test case for :
- Diffie-Hellman  key pair generation and shared secret compute
- DSA sign and verify

Test cases uses predefined test vectors.

Signed-off-by: Sunila Sahu <sunila.sahu@caviumnetworks.com>
Signed-off-by: Shally Verma <shally.verma@caviumnetworks.com>
Signed-off-by: Ashish Gupta <ashish.gupta@caviumnetworks.com>
Signed-off-by: Umesh Kartha <umesh.kartha@caviumnetworks.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

test/crypto: add RSA and Mod tests

Test application include test case for :
- RSA encrypt, decrypt, sign and verify
- Modular Inversion and Exponentiation

Test cases uses predefined test vectors.

Signed-off-by: Sunila Sahu <sunila.sahu@caviumnetworks.com>
Signed-off-by: Shally Verma <shally.verma@caviumnetworks.com>
Signed-off-by: Ashish Gupta <ashish.gupta@caviumnetworks.com>
Signed-off-by: Umesh Kartha <umesh.kartha@caviumnetworks.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

doc: add asymmetric crypto in programmer guide

Update cryptodev programmer guide with description of
asymmetric crypto framework in lib cryptodev.

Signed-off-by: Shally Verma <shally.verma@caviumnetworks.com>
Signed-off-by: Sunila Sahu <sunila.sahu@caviumnetworks.com>
Signed-off-by: Ashish Gupta <ashish.gupta@caviumnetworks.com>
Signed-off-by: Umesh Kartha <umesh.kartha@caviumnetworks.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

cryptodev: add asymmetric capability

Extend cryptodev with asymmetric capability APIs and
definitions.

Signed-off-by: Shally Verma <shally.verma@caviumnetworks.com>
Signed-off-by: Sunila Sahu <sunila.sahu@caviumnetworks.com>
Signed-off-by: Ashish Gupta <ashish.gupta@caviumnetworks.com>
Signed-off-by: Umesh Kartha <umesh.kartha@caviumnetworks.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

cryptodev: support asymmetric operations

Extend DPDK librte_cryptodev to:
- define asym op type in rte_crypto_op_type and associated
  op pool create/alloc APIs
- define asym session and associated session APIs

If PMD shows in its feature flag that it supports both sym and
asym then it must support those on all its qps.

Signed-off-by: Shally Verma <shally.verma@caviumnetworks.com>
Signed-off-by: Sunila Sahu <sunila.sahu@caviumnetworks.com>
Signed-off-by: Ashish Gupta <ashish.gupta@caviumnetworks.com>
Signed-off-by: Umesh Kartha <umesh.kartha@caviumnetworks.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

cryptodev: add asymmetric xform and op definitions

Add rte_crypto_asym.h with supported xforms
and associated op structures and APIs

API currently supports:
- RSA Encrypt, Decrypt, Sign and Verify
- Modular Exponentiation and Inversion
- DSA Sign and Verify
- Diffie-Hellman private key exchange
- Diffie-Hellman public key exchange
- Diffie-Hellman shared secret compute
- Diffie-Hellman public/private key pair generation
using xform chain

Signed-off-by: Shally Verma <shally.verma@caviumnetworks.com>
Signed-off-by: Sunila Sahu <sunila.sahu@caviumnetworks.com>
Signed-off-by: Ashish Gupta <ashish.gupta@caviumnetworks.com>
Signed-off-by: Umesh Kartha <umesh.kartha@caviumnetworks.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

crypto/scheduler: add minimum head/tailroom requirement

Minimum head/tailroom requirement for each PMD has to be considered
while populating the dev_info.

Signed-off-by: Anoob Joseph <anoob.joseph@caviumnetworks.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

test/crypto: skip validation of head/tailroom used by PMD

Crypto PMDs would specify the head/tailroom it would use while
processing the crypto requests. This need to be considered while
verifying buffers processed by crypto PMDs.

Signed-off-by: Anoob Joseph <anoob.joseph@caviumnetworks.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

app/crypto-perf: honour min headroom/tailroom

Crypto dev would specify its headroom and tailroom requirement and the
application is expected to honour this while creating buffers.

Signed-off-by: Anoob Joseph <anoob.joseph@caviumnetworks.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

cryptodev: add min headroom and tailroom requirement

Enabling crypto devs to specify the minimum headroom and tailroom it
expects in the mbuf. For net PMDs, standard headroom has to be honoured
by applications, which is not strictly followed for crypto devs. This
prevents crypto devs from using free space in mbuf (available as
head/tailroom) for internal requirements in crypto operations. Addition
of head/tailroom requirement will help PMDs to communicate such
requirements to the application.

The availability and use of head/tailroom is an optimization if the
hardware supports use of head/tailroom for crypto-op info. For devices
that do not support using the head/tailroom, they can continue to operate
without any performance-drop.

Signed-off-by: Anoob Joseph <anoob.joseph@caviumnetworks.com>
Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

cryptodev: rename experimental private data APIs

The name private_data is confusing in these APIs:
rte_cryptodev_sym_session_set_private_data()
rte_cryptodev_sym_session_get_private_data()
It refers to data added at the end of the session hdr for
use by the application.
The session already contains sess_private_data[index]
which is used to store private pmd data and most references to private
data refer to that.
e.g. external apis
rte_cryptodev_sym_get_private_session_size() and internal
set/get_session_private_data() refer to sess_private_data[].

So rename to user_data, i.e.
rte_cryptodev_sym_session_set_user_data()
rte_cryptodev_sym_session_get_user_data()

Refers to changes introduced here:
https://patches.dpdk.org/patch/38172/

Signed-off-by: Fiona Trahe <fiona.trahe@intel.com>
Acked-by: Abhinandan Gujjar <abhinandan.gujjar@intel.com>

cryptodev: check if symmetric sessions are supported

Since asymmetric functionality will be implemented soon,
not all PMDs must support symmetric sessions.
Therefore, a check is added if a device does not implement
the symmetric functions, meaning that the device does not
support symmetric operations.

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Acked-by: Akhil Goyal <akhil.goyal@nxp.com>

cryptodev: rename PMD symmetric session API

The PMD specific API to configure, clear and
obtain session private size is renamed, including
the word _sym_ to clarify that it is API
for symmetric sessions, so there will not be any
conflicts for asymmetric and other type of sessions
in the future.

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Acked-by: Akhil Goyal <akhil.goyal@nxp.com>

cryptodev: remove attach/detach session API

As announced in the previous release,
The API to attach/dettach a session to a queue pair
is removed, as it was only used in DPAA, and it is not
actually needed.

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Acked-by: Akhil Goyal <akhil.goyal@nxp.com>

cryptodev: replace mbuf scatter gather flag

The current mbuf scatter gatter feature flag is
too ambiguous, as it is not clear if input and/or output
buffers can be scatter gather mbufs or not, plus
if in-place and/or out-of-place is supported.

Therefore, five new flags will replace this flag:
- RTE_CRYPTODEV_FF_IN_PLACE_SGL
- RTE_CRYPTODEV_FF_OOP_SGL_IN_SGL_OUT
- RTE_CRYPTODEV_FF_OOP_SGL_IN_LB_OUT
- RTE_CRYPTODEV_FF_OOP_LB_IN_SGL_OUT
- RTE_CRYPTODEV_FF_OOP_LB_IN_LB_OUT

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Acked-by: Akhil Goyal <akhil.goyal@nxp.com>

cryptodev: remove old get session size functions

Removed rte_cryptodev_get_header_session_size
and rte_cryptodev_get_private_session_size functions,
as they have been substituted with functions
specific for symmetric operations, with _sym_ word
after "rte_cryptodev_".

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Acked-by: Akhil Goyal <akhil.goyal@nxp.com>

cryptodev: remove queue start/stop functions

Removed cryptodev queue start/stop functions,
as they were marked deprecated in 18.05, since they
were not implemented by any driver.

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Acked-by: Akhil Goyal <akhil.goyal@nxp.com>

doc: remove unneeded deprecation notice

In release 18.05, a deprecation notice to remove the `sym`
structure in the cryptodev info structure was sent.
However, only one of the fields inside the structure will
be removed, so the notice is not actually correct.
In any case, it needs to be removed.

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Acked-by: Akhil Goyal <akhil.goyal@nxp.com>

cryptodev: remove max number of sessions parameter

Most crypto PMDs do not have a limitation
of the number of the sessions that can be handled
internally. The value that was set before was not
actually used at all, since the sessions are created
at the application level.
Therefore, this value is not parsed from the initial
crypto parameters anymore and it is set to 0,
meaning that there is no actual limit.

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Acked-by: Akhil Goyal <akhil.goyal@nxp.com>

cryptodev: define value for unlimited sessions

Currently, the info structure contains the maximum number
of sessions that a device can manage.
This field was useful when the session mempool was created inside
each device, but now it is created at the application level.

Most PMDs do not have a limitation on the sessions managed,
but a few do, therefore this field must remain in the structure.
However, a new value, 0, can be used to indicate that
a device does not have an actual maximum of sessions.

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Acked-by: Akhil Goyal <akhil.goyal@nxp.com>

crypto/mvsam: parse max number of sessions

The maximum number of sessions device argument will be removed,
as most PMDs do not have a limitation on this number.
Therefore, the MVSAM PMD needs to parse this value internally.

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Acked-by: Tomasz Duszynski <tdu@semihalf.com>

examples/ipsec-secgw: check for max supported sessions

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Acked-by: Akhil Goyal <akhil.goyal@nxp.com>

examples/l2fwd-crypto: limit number of sessions

Calculate the number of sessions required for the application,
knowing that there is only one session required per device.

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Acked-by: Akhil Goyal <akhil.goyal@nxp.com>

test/crypto: limit number of sessions

Instead of using the maximum number of sessions
allowed by the PMDs (which will change to unlimited most
of the PMDs), limit the number to a small sufficient amount.

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Acked-by: Akhil Goyal <akhil.goyal@nxp.com>

app/crypto-perf: limit number of sessions

Instead of creating a fixed number of sessions,
calculate the necessary number based on number of devices
and queue pairs used.

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Acked-by: Akhil Goyal <akhil.goyal@nxp.com>

cryptodev: remove max number of sessions per queue

The cryptodev info structure currently contains
the maximum number of sessions that can be used in a queue pair.
This is only set in DPAA_SEC PMD, and since it is calculated
based on the maximum number of sessions (which is not used
anymore), this field can be removed.

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Acked-by: Akhil Goyal <akhil.goyal@nxp.com>

cryptodev: replace bus specific struct with generic dev

Structure rte_cryptodev_info has currently PCI device
information ("struct rte_pci_device") in it.

This information is not generic to all devices,
so this gets replaced with the generic "rte_device" structure,
compatible with all crypto devices.

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Acked-by: Akhil Goyal <akhil.goyal@nxp.com>

compressdev: add huffman encoding flags

Added Huffman fixed and dynamic encoding feature flags,
so an application can query if a device supports
these two types, when performing DEFLATE compression.

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

compressdev: replace mbuf scatter gather flag

The current mbuf scatter gather feature flag is
too ambiguous, as it is not clear if input and/or output
buffers can be scatter gather mbufs or not.

Therefore, three new flags will replace this flag:
- RTE_COMP_FF_OOP_SGL_IN_SGL_OUT
- RTE_COMP_FF_OOP_SGL_IN_FB_OUT
- RTE_COMP_FF_OOP_LB_IN_SGL_OUT

Note that out-of-place flat buffers is supported by default
and in-place is not supported by the library.

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

doc: rename compress feature flag

Renamed feature "Bypass" to "Pass-through",
as it is a more explicit name, meaning that the PMD
is capable of passing the mbufs through it,
without making any modifications (i.e.. NULL algorithm).

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

doc: cleanup ISA-L PMD feature matrix

In PMD feature matrices (.ini files), it is not required to
have the list of features that are not supported,
just the ones that are.

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Acked-by: Lee Daly <lee.daly@intel.com>