Context memory is allocated from mempool. Ideally
it will get all memory from single segment, so simple offset
calculation is used for address conversion for such addresses
from context memory.
Akhil Goyal [Thu, 14 Dec 2017 06:52:02 +0000 (12:22 +0530)]
examples/ipsec-secgw: add cryptodev mask option
Previously, ipsec-secgw application did not give user the
flexibility to decide which crypto device(s) will be used.
In this patch, a new cryptodev_mask option is added to the
application. Same as portmask, the cryptodev_mask avails the
user to mask out the unwanted crypto devices in the system.
This patch is similar to the support added in l2fwd-crypto
(d2797f51cc63: examples/l2fwd-crypto: add cryptodev mask option)
Signed-off-by: Akhil Goyal <akhil.goyal@nxp.com> Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Anoob Joseph [Mon, 18 Dec 2017 07:15:05 +0000 (07:15 +0000)]
examples/ipsec-secgw: support inline protocol
Adding support for inline protocol processing
In ingress side, application will receive regular IP packets, without
any IPsec related info. Application will do a selector check (SP-SA
check) by making use of the metadata from the packet. The
device-specific metadata in mbuf would aid in determing the security
session which processed the packet.
In egress side, the plain packet would be submitted to the driver. The
packet will have optional metadata, which could be used to identify the
security session associated with the packet.
Signed-off-by: Anoob Joseph <anoob.joseph@caviumnetworks.com> Reviewed-by: Akhil Goyal <akhil.goyal@nxp.com>
Anoob Joseph [Mon, 18 Dec 2017 07:15:04 +0000 (07:15 +0000)]
security: support user data retrieval
In case of inline protocol processed ingress traffic, the packet may not
have enough information to determine the security parameters with which
the packet was processed. In such cases, application could get metadata
from the packet which could be used to identify the security parameters
with which the packet was processed.
Application could register "userdata" with the security session, and
this could be retrieved from the metadata of inline processed packets.
The metadata returned by "rte_security_get_pkt_metadata()" will be
device specific. Also the driver is expected to return the application
registered "userdata" as is, without any modifications.
Signed-off-by: Anoob Joseph <anoob.joseph@caviumnetworks.com> Acked-by: Akhil Goyal <akhil.goyal@nxp.com>
Pablo de Lara [Wed, 13 Dec 2017 13:14:08 +0000 (13:14 +0000)]
app/crypto-perf: support IMIX
Add support for IMIX performance tests, where a distribution
of various packet sizes can be submitted to a crypto
device, testing a closer to a real world scenario.
A sequence of packet sizes, selected randomly from a list of packet
sizes (with "buffer-sz" parameter) with a list of the weights
per packet size (using "imix" parameter), is generated
(the length of this sequence is the same length as the pool,
set with "pool-sz" parameter).
This sequence is used repeteadly for all the crypto
operations submitted to the crypto device (with "--total-ops" parameter).
Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com> Acked-by: Fan Zhang <roy.fan.zhang@intel.com>
Pablo de Lara [Wed, 13 Dec 2017 15:55:00 +0000 (15:55 +0000)]
cryptodev: fix function prototype
qp_detach_session function was using the attach_session_t
function prototype, instead of detach_session_t.
Since both of them have the same parameters, there were
no compilation issues, but it is not consistent.
Fixes: d816fdea557c ("cryptodev: add API to associate session with queue pair") Cc: stable@dpdk.org Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Nélio Laranjeiro [Thu, 23 Nov 2017 10:02:39 +0000 (11:02 +0100)]
security: fix pedantic compilation
/x86_64-native-linuxapp-gcc/include/rte_security.h:229:8:
error: struct has no members [-Werror=pedantic]
struct rte_security_macsec_xform {
^~~~~~~~~~~~~~~~~~~~~~~~~
/x86_64-native-linuxapp-gcc/include/rte_security.h:453:3:
error: struct has no members [-Werror=pedantic]
struct {
^~~~~~
Fixes: c261d1431bd8 ("security: introduce security API and framework") Cc: stable@dpdk.org Signed-off-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com> Acked-by: Akhil Goyal <akhil.goyal@nxp.com>
Nélio Laranjeiro [Thu, 23 Nov 2017 10:02:38 +0000 (11:02 +0100)]
crypto: fix pedantic compilation
/x86_64-native-linuxapp-gcc/include/rte_crypto.h:126:28:
error: ISO C forbids zero-size array ‘sym’ [-Werror=pedantic]
struct rte_crypto_sym_op sym[0];
^~~
Zero-size array is an extension to the language it cannot be replaced by a
empty size array i.e. [] because structure is inside a union.
Fixes: d2a4223c4c6d ("cryptodev: do not store pointer to op specific params") Cc: stable@dpdk.org Signed-off-by: Nelio Laranjeiro <nelio.laranjeiro@6wind.com> Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Anoob Joseph [Tue, 12 Dec 2017 08:32:25 +0000 (08:32 +0000)]
examples/ipsec-secgw: fix usage of incorrect port
When security offload is enabled, the packet should be forwarded on the
port configured in the SA. Security session will be configured on that
port only, and sending the packet on other ports could result in
unencrypted packets being sent out.
This would have performance improvements too, as the per packet LPM
lookup would be avoided for IPsec packets, in inline mode.
Fixes: ec17993a145a ("examples/ipsec-secgw: support security offload") Signed-off-by: Anoob Joseph <anoob.joseph@caviumnetworks.com> Acked-by: Akhil Goyal <akhil.goyal@nxp.com> Acked-by: Radu Nicolau <radu.nicolau@intel.com>
Andrea Grandi [Wed, 22 Nov 2017 18:03:12 +0000 (10:03 -0800)]
doc: fix format in OpenSSL installation guide
List of supported OpenSSL versions and code block with dependencies were
not properly formatted.
Fixes: d61f70b4c918 ("crypto/libcrypto: add driver for OpenSSL library") Cc: stable@dpdk.org Signed-off-by: Andrea Grandi <andrea.grandi@intel.com> Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Andrea Grandi [Wed, 22 Nov 2017 18:03:11 +0000 (10:03 -0800)]
doc: fix lists of supported crypto algorithms
Add a missing space must before the first item of the list to display it
correctly in the User Guide.
Fixes: d61f70b4c918 ("crypto/libcrypto: add driver for OpenSSL library") Fixes: b79e4c00af0e ("cryptodev: use AES-GCM/CCM as AEAD algorithms") Cc: stable@dpdk.org Signed-off-by: Andrea Grandi <andrea.grandi@intel.com> Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Aleksey Baulin [Sun, 19 Nov 2017 22:16:04 +0000 (01:16 +0300)]
eal: improve likely and unlikely macros
A warning is issued when using an argument to likely() or unlikely()
builtins which is evaluated to a pointer value, as __builtin_expect()
expects a 'long int' type for its first argument. With this fix
a pointer value is converted to an integer with the value of 0 or 1.
Herbert Guan [Fri, 19 Jan 2018 06:10:36 +0000 (14:10 +0800)]
eal/arm64: optimize memcpy
This patch provides an option to do rte_memcpy() using 'restrict'
qualifier, which can induce GCC to do optimizations by using more
efficient instructions, providing some performance gain over memcpy()
on some ARM64 platforms/enviroments.
The memory copy performance differs between different ARM64
platforms. And a more recent glibc (e.g. 2.23 or later)
can provide a better memcpy() performance compared to old glibc
versions. It's always suggested to use a more recent glibc if
possible, from which the entire system can get benefit. If for some
reason an old glibc has to be used, this patch is provided for an
alternative.
This implementation can improve memory copy on some ARM64
platforms, when an old glibc (e.g. 2.19, 2.17...) is being used.
It is disabled by default and needs "RTE_ARCH_ARM64_MEMCPY"
defined to activate. It's not always proving better performance
than memcpy() so users need to run DPDK unit test
"memcpy_perf_autotest" and customize parameters in "customization
section" in rte_memcpy_64.h for best performance.
Compiler version will also impact the rte_memcpy() performance.
It's observed on some platforms and with the same code, GCC 7.2.0
compiled binary can provide better performance than GCC 4.8.5. It's
suggested to use GCC 5.4.0 or later.
Signed-off-by: Herbert Guan <herbert.guan@arm.com> Acked-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>
Maxime Coquelin [Fri, 12 Jan 2018 10:22:20 +0000 (11:22 +0100)]
bus/pci: forbid IOVA mode if IOMMU address width too small
Intel VT-d supports different address widths for the IOVAs, from
39 bits to 56 bits.
While recent processors support at least 48 bits, VT-d emulation
currently only supports 39 bits. It makes DMA mapping to fail in this
case when using VA as IOVA mode, as user-space virtual addresses uses
up to 47 bits (see kernel's Documentation/x86/x86_64/mm.txt).
This patch parses VT-d CAP register value available in sysfs, and
forbid VA as IOVA mode if the GAW is 39 bits or unknown.
Fixes: f37dfab21c98 ("drivers/net: enable IOVA mode for Intel PMDs") Cc: stable@dpdk.org Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com> Tested-by: Chas Williams <chas3@att.com>
Elza Mathew [Mon, 6 Nov 2017 18:04:49 +0000 (10:04 -0800)]
hash: select fbk function at run-time
Compile-time function selection can potentially lead to
lower performance on generic builds done by distros.
Replaced compile time flag checks with run-time function
selection.
Signed-off-by: Elza Mathew <elza.mathew@intel.com> Acked-by: Bruce Richardson <bruce.richardson@intel.com>
Elza Mathew [Mon, 6 Nov 2017 18:04:02 +0000 (10:04 -0800)]
hash: select cuckoo function at run-time
Compile-time function selection can potentially lead to
lower performance on generic builds done by distros.
Replaced compile time flag checks with run-time function
selection.
Signed-off-by: Elza Mathew <elza.mathew@intel.com> Acked-by: Bruce Richardson <bruce.richardson@intel.com>
Nipun Gupta [Wed, 3 Jan 2018 16:44:19 +0000 (22:14 +0530)]
event/dpaa2: return correct number of event ports
In the info get API we were returning number of event
ports as MAX number of LCORE's. After this change actual
number of event ports (i.e. number of DPIO's) is provided
in the info get API.
Sunil Kumar Kori [Tue, 16 Jan 2018 20:43:57 +0000 (02:13 +0530)]
net/dpaa: support Rx queue configurations with eventdev
Given ethernet Rx queues can be attached with event queue in
parallel or atomic mode. Patch imlmplements Rx queue
configuration, attachment/detachment with given event queue and their
corresponding callbacks to handle events from respective queues.
Sunil Kumar Kori [Tue, 16 Jan 2018 20:43:55 +0000 (02:13 +0530)]
bus/dpaa: support event dequeue and consumption
To receive events from given event port, corresponding
function needs to be added which receives events
from portal. Also added function to consume received
events based on entry index.
Signed-off-by: Sunil Kumar Kori <sunil.kori@nxp.com> Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Nipun Gupta [Wed, 17 Jan 2018 11:39:13 +0000 (17:09 +0530)]
bus/fslmc: add flag to configure DCA in QBMAN multi Tx
With the current QBMAN multi-tx API, we need to create separate
enqueue descriptors for each of the packet which is required to
be enqueued to the hardware, once we support Atomic Queues
(with DCA) in dpaa2 drivers. Creating enqueue descriptor for
each packet is costly and have significant performance impact.
This patch introduces a flag parameter in the QBMAN multi-tx API,
so that DCA configuration (and later on ORP/ODP for ordered queues)
can be passed using flags and be updated in the EQCR using this flag.
Pavan Nikhilesh [Tue, 16 Jan 2018 17:46:03 +0000 (23:16 +0530)]
app/eventdev: add pipeline atq test
This is a pipeline test case that aims at testing the following with
``all types queue`` eventdev scheme.
1. Measure the end-to-end performance of an event dev with a ethernet dev.
2. Maintain packet ordering from Rx to Tx.
The atq queue test functions as same as ``pipeline_queue`` test.
The difference is, It uses, ``all type queue scheme`` instead of separate
queues for each stage and thus reduces the number of queues required to
realize the use case.
Note: The --prod_type_ethdev is mandatory for running the application.
Example command to run pipeline atq test:
sudo build/app/dpdk-test-eventdev -c 0xf -s 0x8 --vdev=event_sw0 -- \
--test=pipeline_atq --wlcore=1 --prod_type_ethdev --stlist=ao
Signed-off-by: Pavan Nikhilesh <pbhagavatula@caviumnetworks.com> Acked-by: Harry van Haaren <harry.van.haaren@intel.com>
Pavan Nikhilesh [Tue, 16 Jan 2018 17:46:01 +0000 (23:16 +0530)]
app/eventdev: add pipeline queue test
This is a pipeline queue test case that aims at testing the following:
1. Measure the end-to-end performance of an event dev with a ethernet dev.
2. Maintain packet ordering from Rx to Tx.
The pipeline queue test configures the eventdev with Q queues and P ports,
where Q is (nb_ethdev * nb_stages) + nb_ethdev and P is nb_workers.
The user can choose the number of workers and number of stages through the
--wlcores and the --stlist application command line arguments respectively.
The probed ethernet devices act as producer(s) for this application.
The ethdevs are configured as event Rx adapters that enables them to
injects events to eventdev based the first stage schedule type list
requested by the user through --stlist the command line argument.
Based on the number of stages to process(selected through --stlist),
the application forwards the event to next upstream queue and when it
reaches last stage in the pipeline if the event type is ATOMIC it is
enqueued onto ethdev Tx queue else to maintain ordering the event type is
set to ATOMIC and enqueued onto the last stage queue.
On packet Tx, application increments the number events processed and print
periodically in one second to get the number of events processed in one
second.
Note: The --prod_type_ethdev is mandatory for running the application.
Example command to run pipeline queue test:
sudo build/app/dpdk-test-eventdev -c 0xf -s 0x8 --vdev=event_sw0 -- \
--test=pipeline_queue --wlcore=1 --prod_type_ethdev --stlist=ao
Signed-off-by: Pavan Nikhilesh <pbhagavatula@caviumnetworks.com> Acked-by: Harry van Haaren <harry.van.haaren@intel.com>
Pavan Nikhilesh [Tue, 16 Jan 2018 17:45:59 +0000 (23:15 +0530)]
app/eventdev: add Tx service setup
Setup one port event port for Tx and link the respective event queue.
Register the Tx function as a service to be called from a service core.
The Tx function dequeues the events from the event queue and transmits
the packet to its respective ethernet port.
Signed-off-by: Pavan Nikhilesh <pbhagavatula@caviumnetworks.com> Acked-by: Harry van Haaren <harry.van.haaren@intel.com>
Pavan Nikhilesh [Wed, 10 Jan 2018 11:10:08 +0000 (16:40 +0530)]
examples/eventdev: add all type queue option
Added configurable option to make queue type as all type queues i.e.
RTE_EVENT_QUEUE_CFG_ALL_TYPES based on event dev capability
RTE_EVENT_DEV_CAP_QUEUE_ALL_TYPES.
This can be enabled by supplying '-a' as a cmdline argument.
Signed-off-by: Pavan Nikhilesh <pbhagavatula@caviumnetworks.com> Acked-by: Harry van Haaren <harry.van.haaren@intel.com>
Pavan Nikhilesh [Wed, 10 Jan 2018 11:10:05 +0000 (16:40 +0530)]
examples/eventdev: modify work cycles
The current work cycles function exchanges source and destination mac
address and also pauses the core for the given cycles.
This patch splits the function into two parts i.e. exchange mac and
pause the cores. The pause cores function is invoked at every stage
where as exchange mac is invoked when packet is transmitted.
Signed-off-by: Pavan Nikhilesh <pbhagavatula@caviumnetworks.com> Acked-by: Harry van Haaren <harry.van.haaren@intel.com>
Pavan Nikhilesh [Wed, 10 Jan 2018 11:10:01 +0000 (16:40 +0530)]
examples/eventdev: add framework for caps based pipeline
Add framework to support capability based pipeline.
Based on the capability of event device and probed ethernet devices the
optimal pipeline configuration can be chosen.
Signed-off-by: Pavan Nikhilesh <pbhagavatula@caviumnetworks.com> Acked-by: Harry van Haaren <harry.van.haaren@intel.com>
Pavan Nikhilesh [Wed, 10 Jan 2018 11:09:59 +0000 (16:39 +0530)]
examples/eventdev: support Rx adapter
Use event Rx adapter for packets Rx instead of explicit producer logic.
Use service run iter function for granular control instead of using
dedicated service lcore.
Signed-off-by: Pavan Nikhilesh <pbhagavatula@caviumnetworks.com> Acked-by: Harry van Haaren <harry.van.haaren@intel.com>
Pavan Nikhilesh [Thu, 11 Jan 2018 10:21:47 +0000 (15:51 +0530)]
eventdev: add API to perform self test
Add API to perform self test on the underlying event device driver.
Signed-off-by: Pavan Nikhilesh <pbhagavatula@caviumnetworks.com> Acked-by: Harry van Haaren <harry.van.haaren@intel.com> Acked-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>
Gage Eads [Tue, 9 Jan 2018 16:19:35 +0000 (10:19 -0600)]
event/sw: remove stale IQ references when reconfigured
This commit fixes a bug in which, when the sw PMD is reconfigured, it would
leave stale IQ chunk pointers in each queue's IQ structure. Now, the PMD
initializes all IQs at eventdev start time and releases all IQ chunk
pointers at eventdev stop time (which has the consequence that any events
in a queue when the eventdev is stopped will be lost). This approach should
be resilient to any reconfiguration done between the stop and start, such
as adding or removing queues.
This commit also fixes two potential issues in iq_chunk.h. iq_init()
now initializes the IQ's count field to 0, and iq_dequeue_burst() sets
iq->head to the appropriate next pointer.
Fixes: dca926ca9faa ("event/sw: use dynamically-sized IQs") Reported-by: Pavan Nikhilesh <pbhagavatula@caviumnetworks.com> Signed-off-by: Gage Eads <gage.eads@intel.com> Reviewed-by: Harry van Haaren <harry.van.haaren@intel.com> Acked-by: Pavan Nikhilesh <pbhagavatula@caviumnetworks.com>
Liang Ma [Tue, 9 Jan 2018 14:19:01 +0000 (14:19 +0000)]
doc: add eventdev OPDL PMD guide
Add the description about opdl pmd
Signed-off-by: Liang Ma <liang.j.ma@intel.com> Signed-off-by: Peter Mccarthy <peter.mccarthy@intel.com> Acked-by: Marko Kovacevic <marko.kovacevic@intel.com>
Liang Ma [Tue, 9 Jan 2018 14:18:56 +0000 (14:18 +0000)]
eventdev: extend capability flags
add new capability flags to express the opdl PMD limitations.
RTE_EVENT_DEV_CAP_NONSEQ_MODE
Event device is capable of operating in none sequential mode. The path
of the event is not necessary to be sequential. Application can change
the path of event at runtime. If the flag is not set, then event each event
will follow a path from queue 0 to queue 1 to queue 2 etc. If the flag is
set, events may be sent to queues in any order. If the flag is not set, the
eventdev will return an error when the application enqueues an event for a
qid which is not the next in the sequence.
RTE_EVENT_DEV_CAP_RUNTIME_PORT_LINK
Event device is capable of configuring the queue/port link at runtime.
If the flag is not set, the eventdev queue/port link is only can be
configured during initialization.
RTE_EVENT_DEV_CAP_MULTIPLE_QUEUE_PORT
Event device is capable of setting up the link between multiple queue
with single port. If the flag is not set, the eventdev can only map a
single queue to each port or map a single queue to many port.
Signed-off-by: Liang Ma <liang.j.ma@intel.com> Signed-off-by: Peter Mccarthy <peter.mccarthy@intel.com> Acked-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>
Liang Ma [Tue, 9 Jan 2018 14:18:51 +0000 (14:18 +0000)]
event/opdl: add PMD main body and helper function
This commit adds a OPDL implementation of the eventdev API. The
implementation here is intended to enable the community to use
the OPDL infrastructure under eventdev API.
The main components of the implementation is three files:
- opdl_evdev.c Creation, configuration, etc
- opdl_evdev_xstats.c helper function to support stats collection
- opdl_evdev.h include the main data structure of opdl
device and all the function prototype
need to be exposed to support eventdev API.
- opdl_evdev_init.c implement all initailization helper function
This commit only adds the implementation, no existing DPDK files
are modified.
Signed-off-by: Liang Ma <liang.j.ma@intel.com> Signed-off-by: Peter Mccarthy <peter.mccarthy@intel.com>
Liang Ma [Tue, 9 Jan 2018 14:18:50 +0000 (14:18 +0000)]
event/opdl: add OPDL ring infrastructure library
OPDL ring is the core infrastructure of OPDL PMD. OPDL ring library
provide the core data structure and core helper function set. The Ring
implements a single ring multi-port/stage pipelined packet distribution
mechanism. This mechanism has the following characteristics:
• No multiple queue cost, therefore, latency is significant reduced.
• Fixed dependencies between queue/ports is more suitable for complex.
fixed pipelines of stateless packet processing (static pipeline).
• Has decentralized distribution (no scheduling core).
• Packets remain in order (no reorder core(s)).
* Update build system to enable compilation.
Signed-off-by: Liang Ma <liang.j.ma@intel.com> Signed-off-by: Peter Mccarthy <peter.mccarthy@intel.com> Reviewed-by: Seán Harte <seanbh@gmail.com>
git.droids-corp.org / dpdk.git / log