dpdk.git
3 years agocommon/cpt: rework pending queue
David George [Fri, 24 Sep 2021 11:20:36 +0000 (16:50 +0530)]
common/cpt: rework pending queue

Replace pending queue with one that allows concurrent single producer and
single consumer. This relaxes the restriction of only allowing a single
lcore to operate on a given queue pair.

Signed-off-by: David George <david.george@sophos.com>
Signed-off-by: Anoob Joseph <anoobj@marvell.com>
3 years agoexamples/ipsec-secgw: fix parsing of flow queue
Anoob Joseph [Thu, 16 Sep 2021 06:07:51 +0000 (11:37 +0530)]
examples/ipsec-secgw: fix parsing of flow queue

Documentation specifies that flow port & queue is provided as,

<...> port 0 queue 0

But code is expecting the same as,

<...> port 0 0

Fix the above to match documentation.

Fixes: 8e693616fcb2 ("examples/ipsec-secgw: enable flow based distribution")
Cc: stable@dpdk.org
Signed-off-by: Anoob Joseph <anoobj@marvell.com>
Acked-by: Akhil Goyal <gakhil@marvell.com>
3 years agotest/crypto-perf: test asymmetric crypto throughput
Kiran Kumar K [Thu, 16 Sep 2021 08:38:38 +0000 (14:08 +0530)]
test/crypto-perf: test asymmetric crypto throughput

Added support for asymmetric crypto perf throughput test.
Only modex is supported for now.

One new optype has been added.
--optype modex

./dpdk-test-crypto-perf -c 0x3 -- --devtype crypto_cn9k --optype modex
 --ptest throughput

Signed-off-by: Kiran Kumar K <kirankumark@marvell.com>
Acked-by: Akhil Goyal <gakhil@marvell.com>
3 years agotest/crypto: fix PDCP short MAC-I case
Fan Zhang [Thu, 16 Sep 2021 13:02:42 +0000 (14:02 +0100)]
test/crypto: fix PDCP short MAC-I case

This patch fixes the PDCP short MAC-I test by removing them
from snow3g and kasumi test suite and move to PDCP test suite.
This is to prevent incorrect failure for crypto device not
support PDCP.

Fixes: c24489e479fd ("test/crypto: support PDCP short MAC-I")

Signed-off-by: Fan Zhang <roy.fan.zhang@intel.com>
Acked-by: Anoob Joseph <anoobj@marvell.com>
3 years agoexamples/ipsec-secgw: update event mode inline path
Nithin Dabilpuram [Wed, 15 Sep 2021 16:30:01 +0000 (22:00 +0530)]
examples/ipsec-secgw: update event mode inline path

Update mbuf.l2_len with L2 header size for outbound
inline processing.

This patch also fixes a bug in arg parsing.

Signed-off-by: Nithin Dabilpuram <ndabilpuram@marvell.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
Acked-by: Akhil Goyal <gakhil@marvell.com>
3 years agosecurity: add option for faster user/meta data access
Nithin Dabilpuram [Wed, 15 Sep 2021 16:30:00 +0000 (22:00 +0530)]
security: add option for faster user/meta data access

Currently rte_security_set_pkt_metadata() and rte_security_get_userdata()
methods to set pkt metadata on Inline outbound and get userdata
after Inline inbound processing is always driver specific callbacks.

For drivers that do not have much to do in the callbacks but just
to update metadata in rte_security dynamic field and get userdata
from rte_security dynamic field, having to just to PMD specific
callback is costly per packet operation. This patch provides
a mechanism to do the same in inline function and avoid function
pointer jump if a driver supports the same.

Signed-off-by: Nithin Dabilpuram <ndabilpuram@marvell.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
Acked-by: Akhil Goyal <gakhil@marvell.com>
3 years agombuf: enforce semantics for Tx inline IPsec processing
Nithin Dabilpuram [Wed, 15 Sep 2021 16:29:59 +0000 (21:59 +0530)]
mbuf: enforce semantics for Tx inline IPsec processing

Not all net PMD's/HW can parse packet and identify L2 header and
L3 header locations on Tx. This is inline with other Tx offloads
requirements such as L3 checksum, L4 checksum offload, etc,
where mbuf.l2_len, mbuf.l3_len etc, needs to be set for HW to be
able to generate checksum. Since Inline IPsec is also such a Tx
offload, some PMD's at least need mbuf.l2_len to be valid to
find L3 header and perform Outbound IPSec processing.

Hence, this patch updates documentation to enforce setting
mbuf.l2_len while setting PKT_TX_SEC_OFFLOAD in mbuf.ol_flags
for Inline IPsec Crypto / Protocol offload processing to
work on Tx.

Signed-off-by: Nithin Dabilpuram <ndabilpuram@marvell.com>
Acked-by: Konstantin Ananyev <konstantin.ananyev@intel.com>
Acked-by: Akhil Goyal <gakhil@marvell.com>
3 years agoeal: add macro to swap two variables
Shijith Thotton [Wed, 28 Jul 2021 15:21:42 +0000 (20:51 +0530)]
eal: add macro to swap two variables

Add a macro to swap two variables
and updat common autotest for the same.

Signed-off-by: Shijith Thotton <sthotton@marvell.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
3 years agostack: fix reload head when pop fails
Julien Meunier [Tue, 21 Sep 2021 16:17:24 +0000 (18:17 +0200)]
stack: fix reload head when pop fails

The previous commit 18effad9cfa7 ("stack: reload head when pop fails")
only changed C11 implementation, not generic implementation.

List head must be loaded right before continue (when failed to find the
new head). Without this, one thread might keep trying and failing to pop
items without ever loading the new correct head.

Fixes: 3340202f5954 ("stack: add lock-free implementation")
Cc: stable@dpdk.org
Signed-off-by: Julien Meunier <julien.meunier@nokia.com>
Acked-by: Olivier Matz <olivier.matz@6wind.com>
3 years agovdpa/mlx5: fix large VM memory region registration
Xueming Li [Thu, 23 Sep 2021 08:11:22 +0000 (16:11 +0800)]
vdpa/mlx5: fix large VM memory region registration

When VM size is larger than 4G (u32) and memory region is larger than 4G,
the 32-bit GCD function overflowed and returned wrong value
that resulted in memory registration failure.

This patch calls 64-bit GCD function to avoid overflow.

Fixes: cc07a42da250 ("vdpa/mlx5: prepare memory regions")
Cc: stable@dpdk.org
Signed-off-by: Xueming Li <xuemingl@nvidia.com>
Reviewed-by: Matan Azrad <matan@nvidia.com>
3 years agosched: get 64-bit greatest common divisor
Xueming Li [Thu, 23 Sep 2021 08:11:21 +0000 (16:11 +0800)]
sched: get 64-bit greatest common divisor

This patch adds new function that compute the greatest common
divisor of 64 bits, also changes the original 32 bits function
to call this new 64-bit version.

Cc: stable@dpdk.org
Signed-off-by: Xueming Li <xuemingl@nvidia.com>
Acked-by: Kevin Traynor <ktraynor@redhat.com>
3 years agopipeline: improve handling of learner action arguments
Cristian Dumitrescu [Tue, 14 Sep 2021 19:00:05 +0000 (20:00 +0100)]
pipeline: improve handling of learner action arguments

The arguments of actions that are learned are now specified as part of
the learn instruction as opposed to being statically specified as part
of the learner table configuration.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: enable pipeline compilation
Cristian Dumitrescu [Mon, 13 Sep 2021 16:44:43 +0000 (17:44 +0100)]
pipeline: enable pipeline compilation

Commit the pipeline changes when the compilation process is
successful: change the table lookup instructions to execute the action
function for each action, replace the regular pipeline instructions
with the custom instructions.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: build shared object for pipeline
Cristian Dumitrescu [Mon, 13 Sep 2021 16:44:42 +0000 (17:44 +0100)]
pipeline: build shared object for pipeline

Build the generated C file into a shared object library.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Signed-off-by: Cunming Liang <cunming.liang@intel.com>
3 years agopipeline: generate custom instruction functions
Cristian Dumitrescu [Mon, 13 Sep 2021 16:44:41 +0000 (17:44 +0100)]
pipeline: generate custom instruction functions

Generate a C function for each custom instruction, which essentially
consolidate multiple regular instructions into a single function call.
The pipeline program is split into groups of instructions, and a
custom instruction is generated for each group that has more than one
instruction. Special care is taken the instructions that can do thread
yield (RX, extern) and for those that can change the instruction
pointer (TX, near/far jump).

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: generate action functions
Cristian Dumitrescu [Mon, 13 Sep 2021 16:44:40 +0000 (17:44 +0100)]
pipeline: generate action functions

Generate a C function for each action. For most instructions, the
associated inline function is called directly. Special care is taken
for TX, jump and return instructions.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: export pipeline instructions to file
Cristian Dumitrescu [Mon, 13 Sep 2021 16:44:39 +0000 (17:44 +0100)]
pipeline: export pipeline instructions to file

Export the array of translated instructions to a C file. There is one
such array per action and one for the pipeline.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: introduce pipeline compilation
Cristian Dumitrescu [Mon, 13 Sep 2021 16:44:38 +0000 (17:44 +0100)]
pipeline: introduce pipeline compilation

Lay the foundation to generate C code for the pipeline: C functions
for actions and custom instructions are generated, built as shared
object library and loaded into the pipeline.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: introduce custom instructions
Cristian Dumitrescu [Mon, 13 Sep 2021 16:44:37 +0000 (17:44 +0100)]
pipeline: introduce custom instructions

For better performance, the option to create custom instructions when
the program is translated and add them on-the-fly to the pipeline is
now provided. Multiple regular instructions can now be consolidated
into a single C function optimized by the C compiler directly.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: introduce action functions
Cristian Dumitrescu [Mon, 13 Sep 2021 16:44:36 +0000 (17:44 +0100)]
pipeline: introduce action functions

For better performance, the option to run a single function per action
is now provided, which requires a single function call per action that
can be better optimized by the C compiler, as opposed to one function
call per instruction. Special table lookup instructions are added to
to support this feature.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: enable persistent instruction meta-data
Cristian Dumitrescu [Mon, 13 Sep 2021 16:44:35 +0000 (17:44 +0100)]
pipeline: enable persistent instruction meta-data

Save the instruction meta-data for later use instead of freeing it up
once the instruction translation is completed.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: create inline functions for instruction operands
Cristian Dumitrescu [Mon, 13 Sep 2021 16:44:34 +0000 (17:44 +0100)]
pipeline: create inline functions for instruction operands

Create inline functions to get the instruction operands.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: create inline functions for meter instructions
Cristian Dumitrescu [Mon, 13 Sep 2021 16:44:33 +0000 (17:44 +0100)]
pipeline: create inline functions for meter instructions

Create inline functions for the meter instructions.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: create inline functions for register instructions
Cristian Dumitrescu [Mon, 13 Sep 2021 16:44:32 +0000 (17:44 +0100)]
pipeline: create inline functions for register instructions

Create inline functions for the register instructions.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: create inline functions for ALU instructions
Cristian Dumitrescu [Mon, 13 Sep 2021 16:44:31 +0000 (17:44 +0100)]
pipeline: create inline functions for ALU instructions

Create inline functions for the ALU instructions.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: create inline functions for DMA instruction
Cristian Dumitrescu [Mon, 13 Sep 2021 16:44:30 +0000 (17:44 +0100)]
pipeline: create inline functions for DMA instruction

Create inline functions for the DMA instruction.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: create inline functions for move instruction
Cristian Dumitrescu [Mon, 13 Sep 2021 16:44:29 +0000 (17:44 +0100)]
pipeline: create inline functions for move instruction

Create inline functions for the move instruction.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: create inline functions for extern instruction
Cristian Dumitrescu [Mon, 13 Sep 2021 16:44:28 +0000 (17:44 +0100)]
pipeline: create inline functions for extern instruction

Create inline functions for the extern instruction.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: create inline functions for learn instruction
Cristian Dumitrescu [Mon, 13 Sep 2021 16:44:27 +0000 (17:44 +0100)]
pipeline: create inline functions for learn instruction

Create inline functions for the learn and forget instructions.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: create inline functions for validate instruction
Cristian Dumitrescu [Mon, 13 Sep 2021 16:44:26 +0000 (17:44 +0100)]
pipeline: create inline functions for validate instruction

Create inline functions for the validate and invalidate instructions.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: create inline functions for emit instruction
Cristian Dumitrescu [Mon, 13 Sep 2021 16:44:25 +0000 (17:44 +0100)]
pipeline: create inline functions for emit instruction

Create inline functions for the emit instruction.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: create inline functions for extract instruction
Cristian Dumitrescu [Mon, 13 Sep 2021 16:44:24 +0000 (17:44 +0100)]
pipeline: create inline functions for extract instruction

Create inline functions for the extract instruction.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: create inline functions for Tx instruction
Cristian Dumitrescu [Mon, 13 Sep 2021 16:44:23 +0000 (17:44 +0100)]
pipeline: create inline functions for Tx instruction

Create inline functions for the Tx instruction.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: create inline functions for Rx instruction
Cristian Dumitrescu [Mon, 13 Sep 2021 16:44:22 +0000 (17:44 +0100)]
pipeline: create inline functions for Rx instruction

Create inline functions for the Rx instruction.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: move thread inline functions to header file
Cristian Dumitrescu [Mon, 13 Sep 2021 16:44:21 +0000 (17:44 +0100)]
pipeline: move thread inline functions to header file

Move the thread inline functions to the internal header file.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: move data structures to internal header file
Cristian Dumitrescu [Mon, 13 Sep 2021 16:44:20 +0000 (17:44 +0100)]
pipeline: move data structures to internal header file

Start to consolidate the data structures and inline functions required
by the pipeline instructions into an internal header file.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agoexamples/pipeline: add learner table example
Cristian Dumitrescu [Mon, 20 Sep 2021 15:01:33 +0000 (16:01 +0100)]
examples/pipeline: add learner table example

Added the files to illustrate the learner table usage.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agoexamples/pipeline: support learner tables
Cristian Dumitrescu [Mon, 20 Sep 2021 15:01:32 +0000 (16:01 +0100)]
examples/pipeline: support learner tables

Add application-level support for learner tables.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: support learner tables
Cristian Dumitrescu [Mon, 20 Sep 2021 15:01:31 +0000 (16:01 +0100)]
pipeline: support learner tables

Add pipeline level support for learner tables.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agotable: support learner tables
Cristian Dumitrescu [Mon, 20 Sep 2021 15:01:30 +0000 (16:01 +0100)]
table: support learner tables

A learner table is typically used for learning or connection tracking,
where it allows for the implementation of the "add on miss" scenario:
whenever the lookup key is not found in the table (lookup miss), the
data plane can decide to add this key to the table with a given action
with no control plane intervention. Likewise, the table keys expire
based on a configurable timeout and are automatically deleted from the
table with no control plane intervention.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agoexamples/pipeline: add variable size headers
Cristian Dumitrescu [Tue, 27 Jul 2021 17:43:40 +0000 (18:43 +0100)]
examples/pipeline: add variable size headers

Added the files to illustrate the variable size header usage.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: add header look-ahead instruction
Cristian Dumitrescu [Tue, 27 Jul 2021 17:43:39 +0000 (18:43 +0100)]
pipeline: add header look-ahead instruction

Added look-ahead instruction to read a header from the input packet
without advancing the extraction pointer. This is typically used in
correlation with the special extract instruction to extract variable
size headers from the input packet: the first few header fields are
read without advancing the extraction pointer, just enough to detect
the actual length of the header (e.g. IPv4 IHL field); then the full
header is extracted.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: add variable size headers extract instruction
Cristian Dumitrescu [Tue, 27 Jul 2021 17:43:38 +0000 (18:43 +0100)]
pipeline: add variable size headers extract instruction

Added a mechanism to extract variable size headers through a special
flavor of the extract instruction. The length of the last struct field
which has variable size is passed as argument to the instruction.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: support variable size headers
Cristian Dumitrescu [Tue, 27 Jul 2021 17:43:37 +0000 (18:43 +0100)]
pipeline: support variable size headers

Added support for variable size headers. The last field of a struct
type can now have a variable size between 0 and N bytes. Useful to
accommodate IPv4 packets with options, etc.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agopipeline: prepare for variable size headers
Cristian Dumitrescu [Tue, 27 Jul 2021 17:43:36 +0000 (18:43 +0100)]
pipeline: prepare for variable size headers

The emit instruction that is responsible for pushing headers into the
output packet is now reading the header length from internal run-time
structures as opposed to constant value from the instruction opcode.

Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
3 years agonet/dpaa2: promote some old experimental API
Nipun Gupta [Fri, 3 Sep 2021 07:17:14 +0000 (12:47 +0530)]
net/dpaa2: promote some old experimental API

These APIs were introduced in 19.02, therefore removing
experimental tag to promote them to stable state.

Signed-off-by: Nipun Gupta <nipun.gupta@nxp.com>
3 years agobus/fslmc: move experimental function to internal
Nipun Gupta [Fri, 3 Sep 2021 07:17:13 +0000 (12:47 +0530)]
bus/fslmc: move experimental function to internal

Remove experimental tag from internal API dpaa2_seqn.
This API was introduced in DPDK 20.11 and is now moved to
internal tag.

Signed-off-by: Nipun Gupta <nipun.gupta@nxp.com>
3 years agobus/fslmc: promote experimental VFIO API to stable
Nipun Gupta [Fri, 3 Sep 2021 07:17:12 +0000 (12:47 +0530)]
bus/fslmc: promote experimental VFIO API to stable

This API was introduced in 19.08, therefore removing
experimental tag to promote them to stable state.

Signed-off-by: Nipun Gupta <nipun.gupta@nxp.com>
3 years agobus/dpaa: move experimental function to internal
Nipun Gupta [Fri, 3 Sep 2021 07:17:11 +0000 (12:47 +0530)]
bus/dpaa: move experimental function to internal

Remove experimental tag from internal API dpaa_seqn.
This API was introduced in DPDK 20.11 and is now moved to
internal tag.

Signed-off-by: Nipun Gupta <nipun.gupta@nxp.com>
3 years agometrics: promote deinitialize API
Harman Kalra [Fri, 3 Sep 2021 13:08:03 +0000 (18:38 +0530)]
metrics: promote deinitialize API

Remove experimental flag from rte_metrics_deinit().
This API was introduced in 19.11 release.

Signed-off-by: Harman Kalra <hkalra@marvell.com>
Acked-by: Ray Kinsella <mdr@ashroe.eu>
3 years agoeal/windows: fix debug build
Tal Shnaiderman [Mon, 13 Sep 2021 16:55:00 +0000 (19:55 +0300)]
eal/windows: fix debug build

When building DPDK on Windows in debug mode the following
warning appear:

warning: token pasting of ',' and __VA_ARGS__ is a GNU extension
[-Wgnu-zero-variadic-macro-arguments] #define open(path, flags, ...)
_open(path, flags, ##__VA_ARGS__)

Modify the 'open' macro to avoid it.

Fixes: 45d62067c237 ("eal: make OS shims internal")
Cc: stable@dpdk.org
Signed-off-by: Tal Shnaiderman <talshn@nvidia.com>
Acked-by: Dmitry Kozlyuk <dmitry.kozliuk@gmail.com>
3 years agobus/pci: fix unknown NUMA node value on Windows
Pallavi Kadam [Wed, 22 Sep 2021 21:19:06 +0000 (14:19 -0700)]
bus/pci: fix unknown NUMA node value on Windows

On older CPUs, currently numa_node returns value only for socket 0.
Instead, application should be able to make correct decision and
also to keep consistent with the Linux code,
replace the return value to -1.

Fixes: ac7c98d04f2c ("bus/pci: ignore missing NUMA node on Windows")
Cc: stable@dpdk.org
Reported-by: Vipin Varghese <vipin.varghese@intel.com>
Signed-off-by: Pallavi Kadam <pallavi.kadam@intel.com>
Acked-by: Tal Shnaiderman <talshn@nvidia.com>
3 years agotelemetry: support dict of dicts
Radu Nicolau [Tue, 14 Sep 2021 16:05:16 +0000 (17:05 +0100)]
telemetry: support dict of dicts

Add support for dicts of dicts to telemetry library.
Increase the max string size to 128.

Signed-off-by: Declan Doherty <declan.doherty@intel.com>
Signed-off-by: Radu Nicolau <radu.nicolau@intel.com>
Acked-by: Ciara Power <ciara.power@intel.com>
3 years agodoc: add line continuation note in Meson coding style
Bruce Richardson [Wed, 15 Sep 2021 09:50:36 +0000 (10:50 +0100)]
doc: add line continuation note in Meson coding style

Add a note for the preference of using "()" rather than "\" for line
continuations in Meson.

Suggested-by: David Marchand <david.marchand@redhat.com>
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Reviewed-by: David Marchand <david.marchand@redhat.com>
3 years agodoc: fix numbers power of 2 in LPM6 guide
Ben Pfaff [Mon, 13 Sep 2021 18:46:43 +0000 (11:46 -0700)]
doc: fix numbers power of 2 in LPM6 guide

Fixes: fc1f2750a3ec ("doc: programmers guide")
Cc: stable@dpdk.org
Signed-off-by: Ben Pfaff <blp@ovn.org>
Reviewed-by: David Marchand <david.marchand@redhat.com>
Acked-by: Vladimir Medvedkin <vladimir.medvedkin@intel.com>
3 years agoeal: reword logs for CPU and NUMA counts
Thomas Monjalon [Thu, 16 Sep 2021 10:32:09 +0000 (12:32 +0200)]
eal: reword logs for CPU and NUMA counts

Some logs about cores and nodes were using hypotetic plural (s) form.
A fixed plural form with value at the end is preferred.

Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
3 years agodoc: remove references to the old build system
Thomas Monjalon [Wed, 15 Sep 2021 16:28:45 +0000 (18:28 +0200)]
doc: remove references to the old build system

Some docs and comments in Meson files are still mentioning
the old build system based on "make", removed in 20.11.
After one year, such references are better to be removed.

Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: David Marchand <david.marchand@redhat.com>
3 years agodoc: remove template comments in old release notes
Thomas Monjalon [Wed, 15 Sep 2021 16:25:09 +0000 (18:25 +0200)]
doc: remove template comments in old release notes

The release notes comments mention how to generate the documentation
with the old & removed build system.

Rather than fixing these comments, all old release notes comments
are removed, because they are useful only for the current release.

Few extra blank lines are removed for consistency.

Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: David Marchand <david.marchand@redhat.com>
Acked-by: John McNamara <john.mcnamara@intel.com>
3 years agolib: remove C++ include guard from private headers
Thomas Monjalon [Wed, 15 Sep 2021 16:46:35 +0000 (18:46 +0200)]
lib: remove C++ include guard from private headers

The private headers are compiled internally with a C compiler.
Thus extern "C" declaration is useless in such files.

Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
3 years agomaintainers: sort NXP raw drivers
Thomas Monjalon [Wed, 15 Sep 2021 16:49:30 +0000 (18:49 +0200)]
maintainers: sort NXP raw drivers

Drivers are alphabetically sorted.

Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: David Marchand <david.marchand@redhat.com>
3 years agoregex/mlx5: fix leak after probing failure
Michael Baum [Tue, 31 Aug 2021 20:38:38 +0000 (23:38 +0300)]
regex/mlx5: fix leak after probing failure

In RegEx device probing, there is register read trying after context
device creation.

When the reading fails, the context device was not freed what caused a
memory leak.

Free it.

Fixes: f324162e8e77 ("regex/mlx5: support combined rule file")
Cc: stable@dpdk.org
Signed-off-by: Michael Baum <michaelba@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
Acked-by: Ori Kam <orika@nvidia.com>
3 years agonet/ice/base: support L4 for QinQ switch filter
Steve Yang [Fri, 10 Sep 2021 08:54:58 +0000 (08:54 +0000)]
net/ice/base: support L4 for QinQ switch filter

This patch adds more dummy packet types for QinQ packet,
it enables tcp/udp layer of ipv4/ipv6 for QinQ payload,
so we can use L4 dst/src port as input set for switch
filter.

Signed-off-by: Steve Yang <stevex.yang@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
3 years agonet/ice: support L4 for QinQ switch filter
Steve Yang [Fri, 10 Sep 2021 08:54:57 +0000 (08:54 +0000)]
net/ice: support L4 for QinQ switch filter

Add L4 support for QinQ switch filter as following flow patterns:
eth / vlan / vlan / ipv4 / udp
eth / vlan / vlan / ipv4 / tcp
eth / vlan / vlan / ipv6 / udp
eth / vlan / vlan / ipv6 / tcp

Signed-off-by: Steve Yang <stevex.yang@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
3 years agonet/iavf: fix resource leak on probing failure
Qiming Chen [Fri, 10 Sep 2021 07:48:35 +0000 (15:48 +0800)]
net/iavf: fix resource leak on probing failure

During the port probe process, there are two abnormal branches that did
not release the previously requested memory, resulting in leakage. The
patch adds an iavf_uninit_vf function, which corresponds to the
iavf_init_vf function.

Fixes: ff2d0c345c3b ("net/iavf: support generic flow API")
Cc: stable@dpdk.org
Signed-off-by: Qiming Chen <chenqiming_huawei@163.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
3 years agonet/iavf: fix Rx queue buffer size alignment
Qiming Chen [Sat, 11 Sep 2021 02:07:56 +0000 (10:07 +0800)]
net/iavf: fix Rx queue buffer size alignment

The RTE_ALIGN macro is aligned upwards. If the buf_size variable is not
aligned with 1 << I40E_RXQ_CTX_DBUFF_SHIFT, the rx_buf_len is larger than
the actual mbuf memory after the operation. When receiving the packet, if
the packet is larger than the configured buf_size, it will cause a memory
stepping event.

The patch uses the RTE_ALIGN_FLOOR down alignment macro to correct the
problem.

Fixes: 69dd4c3d0898 ("net/avf: enable queue and device")
Cc: stable@dpdk.org
Signed-off-by: Qiming Chen <chenqiming_huawei@163.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
3 years agonet/i40e/base: fix resource leakage
Qiming Chen [Sat, 21 Aug 2021 06:30:08 +0000 (14:30 +0800)]
net/i40e/base: fix resource leakage

In the i40e_init_arq function, when the i40e_config_arq_regs function
returns from processing failure, the previously applied arq_bufs resource
is not released, which leads to leakage.
The patch is processed in the same way as the i40e_init_asq function,
maintaining a unified coding style.

Fixes: 49ea51605be4 ("net/i40e/base: gracefully clean the resources")
Cc: stable@dpdk.org
Signed-off-by: Qiming Chen <chenqiming_huawei@163.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
3 years agonet/iavf: fix mbuf leak
Qiming Chen [Sat, 11 Sep 2021 01:47:09 +0000 (09:47 +0800)]
net/iavf: fix mbuf leak

A local test found that repeated port start and stop operations during
the continuous SSE vector bufflist receiving process will cause the mbuf
resource to run out. The final positioning is when the port is stopped,
the mbuf of the pkt_first_seg pointer is not released. Resources leak.
The patch scheme is to judge whether the pointer is empty when the port
is stopped, and release the corresponding mbuf if it is not empty.

Fixes: 69dd4c3d0898 ("net/avf: enable queue and device")
Cc: stable@dpdk.org
Signed-off-by: Qiming Chen <chenqiming_huawei@163.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
3 years agonet/ice: fix crash on representor port closing
Dapeng Yu [Tue, 7 Sep 2021 02:00:33 +0000 (10:00 +0800)]
net/ice: fix crash on representor port closing

If DCF representor port is closed after DCF port is closed, there will
be segmentation fault because representor accesses the released resource
of DCF port.

This patch checks if the resource is present before accessing.

Fixes: 5674465a32c8 ("net/ice: add DCF VLAN handling")
Cc: stable@dpdk.org
Signed-off-by: Dapeng Yu <dapengx.yu@intel.com>
Acked-by: Haiyue Wang <haiyue.wang@intel.com>
3 years agonet/ice/base: fix PF ID for DCF
Dapeng Yu [Mon, 13 Sep 2021 02:40:02 +0000 (10:40 +0800)]
net/ice/base: fix PF ID for DCF

In original implementation, if DCF is created on PF1, the PF ID is
still 0, but not 1. Without the right PF ID, the ACL will not work.

This patch makes VF to get its parent's physical function ID.

Fixes: 0b02c9519432 ("net/ice: handle PF initialization by DCF")
Cc: stable@dpdk.org
Signed-off-by: Dapeng Yu <dapengx.yu@intel.com>
Acked-by: Haiyue Wang <haiyue.wang@intel.com>
3 years agonet/i40e: fix device startup resource release
Qiming Chen [Sat, 21 Aug 2021 09:44:35 +0000 (17:44 +0800)]
net/i40e: fix device startup resource release

In the eth_i40e_dev_init function, the tunnel and ethertype hash table
resource release interface should be rte_hash_free instead of rte_free,
and the previously registered interrupt handling function also needs to
be removed from the interrupt list. The patch is amended to use the
correct interface to release the hash table resource and release the
interrupt handling function at the same time.

Fixes: 425c3325f0b0 ("net/i40e: store tunnel filter")
Fixes: 5c53c82c8174 ("net/i40e: store flow director filter")
Cc: stable@dpdk.org
Signed-off-by: Qiming Chen <chenqiming_huawei@163.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
3 years agonet/i40e: fix mbuf leak
Qiming Chen [Mon, 23 Aug 2021 01:50:34 +0000 (09:50 +0800)]
net/i40e: fix mbuf leak

A local test found that repeated port start and stop operations during
the continuous SSE vector bufflist receiving process will cause the mbuf
resource to run out. The final positioning is when the port is stopped,
the mbuf of the pkt_first_seg pointer is not released. Resources leak.
The patch scheme is to judge whether the pointer is empty when the port
is stopped, and release the corresponding mbuf if it is not empty.

Fixes: 4861cde46116 ("i40e: new poll mode driver")
Cc: stable@dpdk.org
Signed-off-by: Qiming Chen <chenqiming_huawei@163.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
3 years agocommon/iavf: update base driver version
Haiyue Wang [Tue, 7 Sep 2021 07:33:04 +0000 (15:33 +0800)]
common/iavf: update base driver version

Update the driver version to trace the change.

Signed-off-by: Haiyue Wang <haiyue.wang@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
3 years agocommon/iavf: remove flow director query opcode
Haiyue Wang [Tue, 7 Sep 2021 07:33:03 +0000 (15:33 +0800)]
common/iavf: remove flow director query opcode

The VIRTCHNL_OP_QUERY_FDIR_FILTER opcode is not used, so remove it.

Signed-off-by: Haiyue Wang <haiyue.wang@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
3 years agocommon/iavf: enable hash calculation based on L4 checksum
Alvin Zhang [Tue, 7 Sep 2021 07:33:02 +0000 (15:33 +0800)]
common/iavf: enable hash calculation based on L4 checksum

Add TCP/UDP/SCTP header checksum field selectors, they can be used in
creating FDIR or RSS rules related to TCP/UDP/SCTP header checksum.

Signed-off-by: Alvin Zhang <alvinx.zhang@intel.com>
Signed-off-by: Haiyue Wang <haiyue.wang@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
3 years agocommon/iavf: add QFI fields for GTPU UL and DL
Junfeng Guo [Tue, 7 Sep 2021 07:33:01 +0000 (15:33 +0800)]
common/iavf: add QFI fields for GTPU UL and DL

The QFI is 6-bit "QoS Flow Identifier" within the GTPU Extension Header.
Add virtchnl fields QFI of GTPU UL/DL for supporting the AVF FDIR.

Signed-off-by: Junfeng Guo <junfeng.guo@intel.com>
Signed-off-by: Haiyue Wang <haiyue.wang@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
3 years agonet/octeontx2: fix MTU when PTP is enabled
Hanumanth Reddy Pothula [Tue, 10 Aug 2021 07:21:00 +0000 (12:51 +0530)]
net/octeontx2: fix MTU when PTP is enabled

Update MTU value based on PTP enable status and reserve eight
bytes in TX path to accommodate VLAN tags.

If PTP is enabled maximum allowed MTU is 9200 otherwise it's 9208.

Fixes: b5dc3140448e ("net/octeontx2: support base PTP")
Cc: stable@dpdk.org
Signed-off-by: Hanumanth Reddy Pothula <hpothula@marvell.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
3 years agonet/cnxk: add callback to get link status
Harman Kalra [Fri, 30 Jul 2021 16:08:29 +0000 (21:38 +0530)]
net/cnxk: add callback to get link status

Adding a new callback for reading the link status. PF can read its
link status and can forward the same to VF once it comes up.

Signed-off-by: Harman Kalra <hkalra@marvell.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
3 years agocommon/cnxk: send link status event to VF
Harman Kalra [Fri, 30 Jul 2021 16:08:28 +0000 (21:38 +0530)]
common/cnxk: send link status event to VF

Currently link event is only sent to the PF by AF as soon as it comes
up, or in case of any physical change in link. PF will broadcast
these link events to all its VFs as soon as it receives it.
But no event is sent when a new VF comes up, hence it will not have
the link status.
Adding support for sending link status to the VF once it comes up
successfully.

Signed-off-by: Harman Kalra <hkalra@marvell.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
3 years agocommon/cnxk: support dual VLAN insert and strip actions
Satheesh Paul [Mon, 19 Jul 2021 05:41:57 +0000 (11:11 +0530)]
common/cnxk: support dual VLAN insert and strip actions

Add ROC API to configure dual VLAN tag addition and removal.

Signed-off-by: Satheesh Paul <psatheesh@marvell.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
3 years agonet/bnxt: fix Rx queue startup state
Lance Richardson [Tue, 14 Sep 2021 12:50:58 +0000 (08:50 -0400)]
net/bnxt: fix Rx queue startup state

Since the addition of support for runtime queue setup,
receive queues that are started by default no longer
have the correct state. Fix this by setting the state
when a port is started.

Fixes: 0105ea1296c9 ("net/bnxt: support runtime queue setup")
Cc: stable@dpdk.org
Signed-off-by: Lance Richardson <lance.richardson@broadcom.com>
Reviewed-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com>
3 years agonet/mlx5: fix RSS expansion traversal over next nodes
Lior Margalit [Sun, 5 Sep 2021 09:35:58 +0000 (12:35 +0300)]
net/mlx5: fix RSS expansion traversal over next nodes

The RSS expansion is based on DFS algorithm to traverse over the possible
expansion paths.

The current implementation breaks out, if it reaches the terminator of
the "next nodes" array, instead of going backwards to try the next path.
For example:
testpmd> flow create 0 ingress pattern eth / ipv6 / udp / vxlan / end
actions rss level 2 types tcp end / end
The paths found are:
ETH IPV6 UDP VXLAN END
ETH IPV6 UDP VXLAN ETH IPV4 TCP END
ETH IPV6 UDP VXLAN ETH IPV6 TCP END
The traversal stopped after getting to the terminator of the next nodes
of the ETH node. It missed the rest of the nodes in the next nodes array
of the VXLAN node.

The fix is to go backwards when reaching the terminator of the current
level and find if there is a "next node" to start traversing a new path.
Using the above example, the flows will be:
ETH IPV6 UDP VXLAN END
ETH IPV6 UDP VXLAN ETH IPV4 TCP END
ETH IPV6 UDP VXLAN ETH IPV6 TCP END
ETH IPV6 UDP VXLAN IPV4 TCP END
ETH IPV6 UDP VXLAN IPV6 TCP END
The traversal will find additional paths, because it traverses through
all the next nodes array of the VXLAN node.

Fixes: 4ed05fcd441b ("ethdev: add flow API to expand RSS flows")
Cc: stable@dpdk.org
Signed-off-by: Lior Margalit <lmargalit@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
3 years agonet/mlx5: fix RSS expansion for explicit graph node
Lior Margalit [Sun, 5 Sep 2021 09:35:57 +0000 (12:35 +0300)]
net/mlx5: fix RSS expansion for explicit graph node

The RSS expansion algorithm is using a graph to find the possible
expansion paths. A graph node with the 'explicit' flag will be skipped,
if it is not found in the flow pattern.

The current implementation misses the case where the node with the
explicit flag is in the middle of the expanded path.
For example:
testpmd> flow create 0 ingress pattern eth / ipv6 / udp / vxlan / end
actions rss level 2 types tcp end / end
The VLAN node has the explicit flag, so it is currently included in the
expanded flow:
ETH IPV6 UDP VXLAN END
ETH IPV6 UDP VXLAN ETH VLAN IPV4 TCP END
ETH IPV6 UDP VXLAN ETH VLAN IPV6 TCP END

The fix is to skip the nodes with the explicit flag while iterating over
the possible expansion paths. Using the above example, the flows will be:
ETH IPV6 UDP VXLAN END
ETH IPV6 UDP VXLAN ETH IPV4 TCP END
ETH IPV6 UDP VXLAN ETH IPV6 TCP END

Fixes: 3f02c7ff6815 ("net/mlx5: fix RSS expansion for inner tunnel VLAN")
Cc: stable@dpdk.org
Signed-off-by: Lior Margalit <lmargalit@nvidia.com>
Acked-by: Matan Azrad <matan@nvidia.com>
3 years agonet/virtio: fix device configure without jumbo Rx offload
Ivan Ilchenko [Thu, 2 Sep 2021 14:39:39 +0000 (17:39 +0300)]
net/virtio: fix device configure without jumbo Rx offload

Use max-pkt-len only if jumbo frames offload is requested
since otherwise this field isn't valid.

Fixes: 8b90e4358112 ("net/virtio: set offload flag for jumbo frames")
Fixes: 4e8169eb0d2d ("net/virtio: fix Rx scatter offload")
Cc: stable@dpdk.org
Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
3 years agovhost: promote some APIs to stable
Chenbo Xia [Tue, 7 Sep 2021 02:58:00 +0000 (10:58 +0800)]
vhost: promote some APIs to stable

As reported by symbol bot, APIs listed in this patch have been
experimental for more than two years. This patch promotes these
18 APIs to stable.

Signed-off-by: Chenbo Xia <chenbo.xia@intel.com>
Acked-by: Ray Kinsella <mdr@ashroe.eu>
Acked-by: Kevin Traynor <ktraynor@redhat.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
3 years agovhost: log socket path on adding connection
Gaoxiang Liu [Tue, 7 Sep 2021 00:51:27 +0000 (08:51 +0800)]
vhost: log socket path on adding connection

Add log print of socket path in vhost_user_add_connection.
It's useful when adding a mass of socket connections,
because the information of every connection is clearer.

Fixes: 8f972312b8f4 ("vhost: support vhost-user")
Cc: stable@dpdk.org
Signed-off-by: Gaoxiang Liu <liugaoxiang@huawei.com>
Reviewed-by: Chenbo Xia <chenbo.xia@intel.com>
3 years agonet/virtio: fix repeated freeing of virtqueue
Gaoxiang Liu [Tue, 31 Aug 2021 14:39:51 +0000 (22:39 +0800)]
net/virtio: fix repeated freeing of virtqueue

When virtio_init_queue returns error, the memory of vq is freed.
But the value of hw->vqs[queue_idx] does not restore.
If virtio_init_queue returns error, the memory of vq is freed again
in virtio_free_queues.

Fixes: 69c80d4ef89b ("net/virtio: allocate queue at init stage")
Cc: stable@dpdk.org
Signed-off-by: Gaoxiang Liu <liugaoxiang@huawei.com>
Reviewed-by: Chenbo Xia <chenbo.xia@intel.com>
3 years agovhost: fix crash on port deletion
Gaoxiang Liu [Thu, 2 Sep 2021 15:45:53 +0000 (23:45 +0800)]
vhost: fix crash on port deletion

The rte_vhost_driver_unregister() and vhost_user_read_cb()
can be called at the same time by 2 threads.
when memory of vsocket is freed in rte_vhost_driver_unregister(),
the invalid memory of vsocket is accessed in vhost_user_read_cb().
It's a bug of both mode for vhost as server or client.

E.g., vhostuser port is created as server.
Thread1 calls rte_vhost_driver_unregister().
Before the listen fd is deleted from poll waiting fds,
"vhost-events" thread then calls vhost_user_server_new_connection(),
then a new conn fd is added in fdset when trying to reconnect.
"vhost-events" thread then calls vhost_user_read_cb() and
accesses invalid memory of socket while thread1 frees the memory of
vsocket.

E.g., vhostuser port is created as client.
Thread1 calls rte_vhost_driver_unregister().
Before vsocket of reconn is deleted from reconn list,
"vhost_reconn" thread then calls vhost_user_add_connection()
then a new conn fd is added in fdset when trying to reconnect.
"vhost-events" thread then calls vhost_user_read_cb() and
accesses invalid memory of socket while thread1 frees the memory of
vsocket.

The fix is to move the "fdset_try_del" in front of free memory of conn,
then avoid the race condition.

The core trace is:
Program terminated with signal 11, Segmentation fault.

Fixes: 52d874dc6705 ("vhost: fix crash on closing in client mode")
Cc: stable@dpdk.org
Signed-off-by: Gaoxiang Liu <liugaoxiang@huawei.com>
Reviewed-by: Chenbo Xia <chenbo.xia@intel.com>
3 years agonet/virtio-user: fix Rx interrupts with multi-queue
David Marchand [Tue, 31 Aug 2021 15:54:11 +0000 (17:54 +0200)]
net/virtio-user: fix Rx interrupts with multi-queue

The callfds[] array stores eventfds sequentially for Rx and Tx vq.

Fixes: 3d4fb6fd2505 ("net/virtio-user: support Rx interrupt")
Cc: stable@dpdk.org
Signed-off-by: David Marchand <david.marchand@redhat.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
3 years agonet/virtio: avoid unneeded link interrupt configuration
David Marchand [Tue, 31 Aug 2021 15:54:10 +0000 (17:54 +0200)]
net/virtio: avoid unneeded link interrupt configuration

There is no reason to re-register a interrupt handler for LSC if this
feature was not requested in the first place.
A simple use case is when asking for Rx interrupts without LSC interrupt.

Fixes: 26b683b4f7d0 ("net/virtio: setup Rx queue interrupts")
Cc: stable@dpdk.org
Signed-off-by: David Marchand <david.marchand@redhat.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
3 years agonet/virtio: report max/min/align Rx descriptor limits
Ivan Ilchenko [Fri, 20 Aug 2021 12:48:12 +0000 (15:48 +0300)]
net/virtio: report max/min/align Rx descriptor limits

Report max/min/align descriptors limits in device info get callback.
Before calling the callback, rte_eth_dev_info_get() provides
default values of nb_min as zero and nb_max as UINT16_MAX that are
not correct for the driver, so one can't rely on them.

Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
3 years agonet/virtio: fix split queue vectorized Rx
Ivan Ilchenko [Fri, 20 Aug 2021 12:47:52 +0000 (15:47 +0300)]
net/virtio: fix split queue vectorized Rx

Descriptors number may be set less than queue size for split queue
vectorized Rx path. Pointers to mbufs for received packets are
obtained from SW ring, that is initially filled with them in the end
of queue setup in virtio_dev_rx_queue_setup_finish(). The begin of the
SW ring filled up to the size of descriptors number. At queue size
offset from the begin of the SW ring pointers to some fake mbuf are also
set for wrapping purpose. So the ring may contains the hole of invalid
pointers from descriptors number offset to queue size offset, and split
vectorized Rx routines could write to the invalid addresses since they
use the ring up to the queue size. Fix this by setting descriptors
number to queue size on Rx queue setup.

Fixes: fc3d66212fed ("virtio: add vector Rx")
Cc: stable@dpdk.org
Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
3 years agonet/virtio: remove handling of zero desc on Rx setup
Ivan Ilchenko [Fri, 20 Aug 2021 12:47:41 +0000 (15:47 +0300)]
net/virtio: remove handling of zero desc on Rx setup

Rx queue setup callback allows to use the whole ring when
descriptor number argument equals zero. There's no point to
handle zero in any way since RTE Rx queue setup function
rte_eth_rx_queue_setup() doesn't pass zero using fallback
values.

Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
3 years agonet/virtio: fix mbuf count on Rx queue setup
Ivan Ilchenko [Fri, 20 Aug 2021 12:47:20 +0000 (15:47 +0300)]
net/virtio: fix mbuf count on Rx queue setup

Rx queue setup finish function may report wrong number of
allocated mbufs in case of in-order feature. Fix the
function to not ignore allocation error and count only
successfully allocated number of buffers.

Fixes: e5f456a98d3c ("net/virtio: support in-order Rx and Tx")
Cc: stable@dpdk.org
Signed-off-by: Ivan Ilchenko <ivan.ilchenko@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
3 years agovhost: remove copy threshold for async path
Jiayu Hu [Fri, 20 Aug 2021 12:44:26 +0000 (08:44 -0400)]
vhost: remove copy threshold for async path

Copy threshold has been introduced in async vhost data
path to select the appropriate copy engine to do copies
for higher efficiency.

However, it may cause packets ordering issues and also
introduces performance unpredictability.

Therefore, this patch removes copy threshold support in
async vhost data path.

Signed-off-by: Jiayu Hu <jiayu.hu@intel.com>
Signed-off-by: Cheng Jiang <cheng1.jiang@intel.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
3 years agonet: fix checksum offload for outer IPv4
Mohsin Kazmi [Tue, 7 Sep 2021 10:49:16 +0000 (12:49 +0200)]
net: fix checksum offload for outer IPv4

Preparation of the headers for the hardware offload
misses the outer IPv4 checksum offload.
It results in bad checksum computed by hardware NIC.

This patch fixes the issue by setting the outer IPv4
checksum field to 0.

Fixes: 4fb7e803eb1a ("ethdev: add Tx preparation")
Cc: stable@dpdk.org
Signed-off-by: Mohsin Kazmi <mohsin.kazmi14@gmail.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
Acked-by: Olivier Matz <olivier.matz@6wind.com>
3 years agoethdev: promote sibling iterators to stable
David Marchand [Mon, 6 Sep 2021 13:02:20 +0000 (15:02 +0200)]
ethdev: promote sibling iterators to stable

This API saw no update since its introduction and will help applications
like OVS ([1] and [2]) that currently look at rte_eth_devices[] to
achieve the same.

1: https://github.com/openvswitch/ovs/blob/master/lib/netdev-dpdk.c#L1285
2: https://github.com/openvswitch/ovs/blob/master/lib/netdev-dpdk.c#L1476

Signed-off-by: David Marchand <david.marchand@redhat.com>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Acked-by: Ray Kinsella <mdr@ashroe.eu>
3 years agoethdev: promote burst mode API
Haiyue Wang [Mon, 6 Sep 2021 05:56:18 +0000 (13:56 +0800)]
ethdev: promote burst mode API

The DPDK Symbol Bot reports:
Please note the symbols listed below have expired. In line with the
DPDK ABI policy, they should be scheduled for removal, in the next
DPDK release.

Symbol
rte_eth_rx_burst_mode_get
rte_eth_tx_burst_mode_get

Signed-off-by: Haiyue Wang <haiyue.wang@intel.com>
Acked-by: Ferruh Yigit <ferruh.yigit@intel.com>
Acked-by: Ray Kinsella <mdr@ashroe.eu>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
3 years agoethdev: fix typo in Rx queue setup API comment
Joyce Kong [Mon, 6 Sep 2021 02:13:53 +0000 (21:13 -0500)]
ethdev: fix typo in Rx queue setup API comment

Fix a typo that mb_pool was misspelt as mp_pool.

Fixes: 4ff702b5dfa9 ("ethdev: introduce Rx buffer split")
Cc: stable@dpdk.org
Signed-off-by: Joyce Kong <joyce.kong@arm.com>
Reviewed-by: Ruifeng Wang <ruifeng.wang@arm.com>
Acked-by: Ferruh Yigit <ferruh.yigit@intel.com>
3 years agoethdev: promote API to set packet types
Pavan Nikhilesh [Thu, 2 Sep 2021 08:17:30 +0000 (13:47 +0530)]
ethdev: promote API to set packet types

Remove experimental tag from rte_eth_dev_set_ptypes().

Signed-off-by: Pavan Nikhilesh <pbhagavatula@marvell.com>
Acked-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru>
Acked-by: Ray Kinsella <mdr@ashroe.eu>
3 years agonet/softnic: promote manage API
Jasvinder Singh [Wed, 1 Sep 2021 12:20:52 +0000 (13:20 +0100)]
net/softnic: promote manage API

This API was introduced in 18.08, therefore removing
experimental tag to promote it to stable state.

Signed-off-by: Jasvinder Singh <jasvinder.singh@intel.com>
Acked-by: Ray Kinsella <mdr@ashroe.eu>