Akhil Goyal [Wed, 27 Mar 2019 11:53:32 +0000 (11:53 +0000)]
crypto/dpaa2_sec: fix offset calculation for GCM
In case of gcm, output buffer should have aad space
before the actual buffer which needs to be written.
CAAM will not write into the aad anything, it will skip
auth_only_len (aad) and write the buffer afterwards.
Fan Zhang [Mon, 25 Mar 2019 13:58:35 +0000 (13:58 +0000)]
crypto/aesni_mb: support newer library version only
As stated in 19.02 deprecation notice, this patch updates the
aesni_mb PMD to remove the support of older Intel-ipsec-mb
library version earlier than 0.52.
Signed-off-by: Fan Zhang <roy.fan.zhang@intel.com> Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Fan Zhang [Mon, 25 Mar 2019 13:51:21 +0000 (13:51 +0000)]
test/crypto: check out of place for AESNI-MB
This patch updates the unit test to enable AESNI-MB PMD
out-of-place tests. A special test type that swap both
the source and destination buffer is added for a more
comprehensive test set to take place.
Signed-off-by: Fan Zhang <roy.fan.zhang@intel.com> Acked-by: Fiona Trahe <fiona.trahe@intel.com>
Arek Kusztal [Thu, 28 Mar 2019 13:37:03 +0000 (14:37 +0100)]
crypto/qat: add modular multiplicative inverse
This commit adds modular multiplicative inverse to Intel
QuickAssist Technology driver. For capabilities or limitations
please refer to qat.rst or qat_asym_capabilities.h.
Arek Kusztal [Thu, 28 Mar 2019 13:37:02 +0000 (14:37 +0100)]
crypto/qat: add modular exponentiation
This commit adds modular exponentiation to Intel QuickAssist
Technology driver. For capabilities or limitations please refer to
qat.rst or qat_asym_capabilities.h.
Anoob Joseph [Thu, 7 Mar 2019 10:39:50 +0000 (10:39 +0000)]
doc: announce ABI change for cryptodev config
Add new field ff_disable in rte_cryptodev_config. This enables
applications to control the features enabled on the crypto device.
Proposed new layout:
/** Crypto device configuration structure */
struct rte_cryptodev_config {
int socket_id; /**< Socket to allocate resources on */
uint16_t nb_queue_pairs;
/**< Number of queue pairs to configure on device */
+ uint64_t ff_disable;
+ /**< Feature flags to be disabled. Only the following features are
+ * allowed to be disabled,
+ * - RTE_CRYPTODEV_FF_SYMMETRIC_CRYPTO
+ * - RTE_CRYPTODEV_FF_ASYMMETRIC_CRYPTO
+ * - RTE_CRYTPODEV_FF_SECURITY
+ */
};
For eth devices, rte_eth_conf.rx_mode.offloads and
rte_eth_conf.tx_mode.offloads fields are used by applications to
control the offloads enabled on the eth device. This proposal adds a
similar ability for the crypto device.
Pavan Nikhilesh [Fri, 29 Mar 2019 07:11:46 +0000 (07:11 +0000)]
app/eventdev: add option for global dequeue timeout
Add option to provide a global dequeue timeout that is used to create
the eventdev.
The dequeue timeout provided will be common across all the worker
ports. If the eventdev hardware supports power management through
dequeue timeout then this option can be used for verifying power
demands at various packet rates.
Harry van Haaren [Wed, 27 Mar 2019 18:45:46 +0000 (18:45 +0000)]
event/sw: fix enqueue checks in self-test
This patch fixes a number of instances of the same return
value mis-check, where previously we checked for a negative
return value as error, however the API returns an unsigned
integer, so these return value checks are invalid.
The rte_event_enqueue_burst() API returns the number of
events enqueued, so in order to identify the error case,
we must check for != the number of intended enqueues.
Fixes: cd1a9e3eab55 ("test/eventdev: add SW tests for load balancing") Cc: stable@dpdk.org Signed-off-by: Harry van Haaren <harry.van.haaren@intel.com>
Pavan Nikhilesh [Sat, 16 Mar 2019 20:27:39 +0000 (20:27 +0000)]
test/event: improve compatibility for timer adapter
Check if eventdev is open system eventdevs i.e. max_num_events = -1
before asserting.
Allow event timer adapter to adjust the resolution using
RTE_EVENT_TIMER_ADAPTER_F_ADJUST_RES and re-calculate timeout ticks
based on the adjusted resolution.
Signed-off-by: Pavan Nikhilesh <pbhagavatula@marvell.com> Acked-by: Erik Gabriel Carrillo <erik.g.carrillo@intel.com>
Jerin Jacob [Wed, 3 Apr 2019 19:28:40 +0000 (19:28 +0000)]
eal: allow to override init macros per OS
baremetal execution environments may have a different
method to enable RTE_INIT instead of using compiler
constructor and/or OS specific linker scheme.
Allow an option to override RTE_INIT* macros using
rte_os.h or appropriate header file.
Signed-off-by: Jerin Jacob <jerinj@marvell.com> Acked-by: Thomas Monjalon <thomas@monjalon.net>
Add unit test to check for hash lookup and bulk-lookup perf for
extendable bucket feature.
It is tested with both lock-free enabled and lock-free disabled case.
Test includes:
- hash lookup on keys in ext bkt
- hash delete causing key-shifts of keys from ext bkt to secondary bkt
Shahaf Shuler [Sun, 31 Mar 2019 08:43:48 +0000 (11:43 +0300)]
mem: limit use of address hint
The commit below added an address hint as starting address for 64-bit
systems in case an explicit base virtual address was not set by the user.
The justification for such hint was to help devices that work in VA
mode and has a address range limitation to work smoothly with the eal
memory subsystem.
While the base address value selected may work fine for the eal
initialization, it easily breaks when trying to register external memory
using rte_extmem_register API.
Trying to register anonymous memory on RH x86_64 machine took several
minutes, during them the function eal_get_virtual_area repeatedly
scanned for a good VA candidate.
The attempt to guess which VA address will be free for mapping will
always result in not portable, error prone code:
* different application may use different libraries along w/ DPDK. One
can never guess which library was called first and how much virtual
memory it consumed.
* external memory can be registered at any time in the application run
time.
In order not to break the existing secondary process design, this patch
only limits the max number of tries that will be done with the
address hint.
When the number of tries exceeds the threshold the code
will use the suggested address from kernel.
Fixes: 1df21702873d ("mem: use address hint for mapping hugepages") Cc: stable@dpdk.org Signed-off-by: Shahaf Shuler <shahafs@mellanox.com> Tested-by: Anatoly Burakov <anatoly.burakov@intel.com> Acked-by: Alejandro Lucero <alejandro.lucero@netronome.com>
Added meson workarounds to build helloworld on Windows.
Windows currently only supports kvargs and eal libraries.
This change restricts the build flow to supported libraries
only.
Added headers to support Windows environment for common source.
These headers will have Windows specific implementions of the
system library APIs provided in Linux and FreeBSD.
Thomas Monjalon [Mon, 1 Apr 2019 23:07:12 +0000 (01:07 +0200)]
eal: remove exec-env directory
Only one header file (rte_kni_common.h) was in the sub-directory
include/exec-env/
This file was installed in a sub-directory of the same name
in the makefile-based build.
Source and install directories are moved as below:
Andrew Rybchenko [Fri, 29 Mar 2019 13:42:14 +0000 (13:42 +0000)]
mbuf: remove Intel offload checks from generic API
rte_validate_tx_offload() is used in Tx prepare callbacks
(RTE_LIBRTE_ETHDEV_DEBUG only) to check Tx offloads consistency.
Requirement that packet headers should not be fragmented is not
documented and unclear where it comes from except
rte_net_intel_cksum_prepare() functions which relies on it.
It could be NIC vendor specific driver or hardware limitation, but,
if so, it should be documented and checked in corresponding Tx
prepare callbacks.
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Olivier Matz <olivier.matz@6wind.com>
Anatoly Burakov [Fri, 29 Mar 2019 17:55:29 +0000 (17:55 +0000)]
mem: do not use lockfiles for single file segments mode
Due to internal glibc limitations [1], DPDK may exhaust internal
file descriptor limits when using smaller page sizes, which results
in inability to use system calls such as select() by user
applications.
Single file segments option stores lock files per page to ensure
that pages are deleted when there are no more users, however this
is not necessary because the processes will be holding onto the
pages anyway because of mmap(). Thus, removing pages from the
filesystem is safe even though they may be used by some other
secondary process. As a result, single file segments mode no
longer stores inordinate amounts of segment fd's, and the above
issue with fd limits is solved.
However, this will not work for legacy mem mode. For that, simply
document that using bigger page sizes is the only option.
On Linux, we currently initialize rte_alarms after
starting to listen for IPC hotplug requests, which gives
us a data race window. Upon receiving such hotplug
request we always try to set an alarm and this obviously
doesn't work if the alarms weren't initialized yet.
To fix it, we initialize alarms before starting to
listen for IPC hotplug messages. Specifically, we move
rte_eal_alarm_init() right after rte_eal_intr_init() as
it makes some sense to keep those two close to each other.
We update the BSD code as well to keep the initialization
order the same in both EAL implementations.
maintainers: claim responsibility for timer library
Add myself as co-maintainer for the timer library.
Signed-off-by: Erik Gabriel Carrillo <erik.g.carrillo@intel.com> Acked-by: Robert Sanford <rsanford@akamai.com> Acked-by: Thomas Monjalon <thomas@monjalon.net>
Bruce Richardson [Thu, 28 Mar 2019 15:30:45 +0000 (15:30 +0000)]
telemetry: fix mapping of statistics
If we have two NIC ports which have a different set of NIC stats we can
end up having two different stats registered with xstats with the same
name. [Since the stats are updated in bulk as a contiguous set, the
second driver re-using the registration of the first is not possible.]
This causes issues with the invalid stat for one driver being found due to
a lookup by name which is unnecessary. Instead of getting stat names
involved do the lookup by ID instead.
Fixes: 1b756087db93 ("telemetry: add parser for client socket messages") Cc: stable@dpdk.org Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: Kevin Laatz <kevin.laatz@intel.com>
David Hunt [Mon, 1 Apr 2019 16:14:41 +0000 (17:14 +0100)]
examples/distributor: detect high frequency cores
The distributor application is bottlenecked by the distributor core,
so if we can give more frequency to this core, then the overall
performance of the application may increase.
This patch uses the rte_power_get_capabilities() API to query the
cores provided in the core mask, and if any high frequency cores are
found (e.g. Turbo Boost is enabled), we will pin the distributor
workload to that core.
Signed-off-by: Liang Ma <liang.j.ma@intel.com> Signed-off-by: David Hunt <david.hunt@intel.com> Reviewed-by: Anatoly Burakov <anatoly.burakov@intel.com>
David Hunt [Mon, 1 Apr 2019 16:14:40 +0000 (17:14 +0100)]
power: add bit for high frequency cores
This patch adds a new bit in the capabilities mask that's returned by
rte_power_get_capabilities(), allowing application to query which cores
have the higher frequencies, and can then pin the workloads accordingly.
David Hunt [Fri, 29 Mar 2019 16:39:20 +0000 (16:39 +0000)]
power: fix governor storage to trim newlines
Currently the Power Libray stores the governor name with an embedded
newline read from the scaling_governor sysfs file. This patch strips
it out.
Fixes: 445c6528b55f ("power: common interface for guest and host") Cc: stable@dpdk.org Signed-off-by: David Hunt <david.hunt@intel.com> Acked-by: Anatoly Burakov <anatoly.burakov@intel.com>
devtools: allow test build outside source directory
The test-meson-builds.sh script correctly detects the source directory and
builds the native builds successfully in a directory outside of the source
tree. However, the paths to the cross-files are not prefixed with the
source directory path, so the cross-builds all fail. Fix this by prepending
the source directory path appropriately.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: Luca Boccassi <bluca@debian.org>
Ali Alnubani [Mon, 1 Apr 2019 15:38:59 +0000 (15:38 +0000)]
build: fix meson build in CI environments
This is to fix a build error with meson in GNU/Linux that is caused
by using the 'more' command to read the VERSION file. The error:
config/meson.build:10:10: ERROR: String
'::::::::::::::\n<RTE_SDK_PATH>VERSION\n::::::::::::::\n19' cannot be
converted to int
The command 'more' prints the file name before the actual
contents of the file when it's being run without a controlling terminal.
This could happen in CI environments.
Please refer to:
https://git.kernel.org/pub/scm/utils/util-linux/util-linux.git/tree/text-utils/more.c
Fixes: c04172b5f031 ("build: add single source of DPDK version number") Fixes: d320fe56bd51 ("build: use version number from config file") Signed-off-by: Ali Alnubani <alialnu@mellanox.com> Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Yongseok Koh [Thu, 28 Mar 2019 18:46:28 +0000 (11:46 -0700)]
app/testpmd: make txonly mode generate multiple flows
Testpmd can generate multiple flows without taking much cost and this
could be a simple traffic generator for developer's quick tests.
If "--txonly-multi-flow" is specified in the command line, IP source
address is varied to generate multiple flows.
Signed-off-by: Yongseok Koh <yskoh@mellanox.com> Acked-by: Bernard Iremonger <bernard.iremonger@intel.com>
Ian Stokes [Fri, 29 Mar 2019 17:52:18 +0000 (17:52 +0000)]
net/e1000: set min and max MTU
This commit sets the min and max supported MTU values for igb devices
via the eth_igb_info_get() function. Min MTU supported is set to
ETHER_MIN_MTU and max MTU is calculated as the max packet length
supported minus the transport overhead. To aid in these calculations
a new MACRO 'E1000_ETH_OVERHEAD' has been introduced to consolidate
overhead calculation and avoid duplication.
Signed-off-by: Ian Stokes <ian.stokes@intel.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
Ian Stokes [Fri, 29 Mar 2019 17:52:17 +0000 (17:52 +0000)]
net/ixgbe: set min and max MTU for VF
This commit sets the min and max supported MTU values for ixgbe VF
devices via the ixgbevf_dev_set_mtu() function. Min MTU supported is
set to ETHER_MIN_MTU and max MTU is calculated as the max packet length
supported minus the transport overhead. As transport overhead is the
same for VF and PF ixgbe devices, reuse MACRO 'IXGBE_ETH_OVERHEAD' to
avoid duplication.
Signed-off-by: Ian Stokes <ian.stokes@intel.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
Ian Stokes [Fri, 29 Mar 2019 17:52:16 +0000 (17:52 +0000)]
net/ixgbe: set min and max MTU
This commit sets the min and max supported MTU values for ixgbe devices
via the ixgbe_dev_info_get() function. Min MTU supported is set to
ETHER_MIN_MTU and max MTU is calculated as the max packet length
supported minus the transport overhead. To aid in these calculations
a new MACRO 'IXGBE_ETH_OVERHEAD' has been introduced to consolidate
overhead calculation and avoid duplication.
Signed-off-by: Ian Stokes <ian.stokes@intel.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
Ian Stokes [Fri, 29 Mar 2019 17:52:15 +0000 (17:52 +0000)]
net/i40e: set min and max MTU for VF
This commit sets the min and max supported MTU values for i40e VF
devices via the i40evf_dev_info_get() function. Min MTU supported
is set to ETHER_MIN_MTU and max MTU is calculated as the max packet
length supported minus the transport overhead.
Signed-off-by: Ian Stokes <ian.stokes@intel.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
Ian Stokes [Fri, 29 Mar 2019 17:52:14 +0000 (17:52 +0000)]
net/i40e: set min and max MTU
This commit sets the min and max supported MTU values for i40e devices
via the i40e_dev_info_get() function. Min MTU supported is set to
ETHER_MIN_MTU and max mtu is calculated as the max packet length
supported minus the transport overhead.
Signed-off-by: Ian Stokes <ian.stokes@intel.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
This addresses the usability issue raised by OVS at DPDK Userspace
summit. It adds general min/max MTU into device info. For compatibility,
and to save space, it fits in a hole in existing structure.
The initial version sets max MTU to normal Ethernet, it is up to
PMD to set larger value if it supports Jumbo frames.
Also remove the deprecation notice introduced in 18.11 regarding this
change and bump ethdev ABI version.
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Signed-off-by: Ian Stokes <ian.stokes@intel.com> Acked-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
net/mlx5: update event handler for multiport IB devices
This patch modifies asynchronous event handler to support multiport
Infiniband devices. Handler queries the event parameters, including
event source port index, and invokes the handler for specific
devices with appropriate port_id.
We are implementing the support for multiport Infiniband device
with representors attached to these multiple ports. Asynchronous
device event notifications (link status change, removal event, etc.)
should be shared between ports. We are going to implement shared
event handler and this patch introduces appropriate device
structure changes and updated event handler install and uninstall
routines.
The code is updated to use the shared IB device context and
device handles. The IB device context is shared between
reprentors created over the single multiport IB device. All
Verbs and DevX objects will be created within this shared context.
net/mlx5: switch to the shared context IB attributes
The code is updated to use the shared IB device attributes,
located in the shared IB context. It saves some memory if
there are representors created over the single Infiniband
device with multiple ports.
The PMD code is updated to use Protected Domain from the
shared IB device context. The Domain is shared between
all devices belonging to the same multiport Infiniband device.
If IB device has only one port, the PD is not shared, because
there is only ethernet device created over IB one.
net/mlx5: switch to the names in the shared IB context
The IB device names are moved from device private data
to the shared context, code involving the names is updated.
The IB port index treatment is added where it is relevant.
The Mellanox NICs support SR-IOV and have E-Switch feature.
When SR-IOV is set up in switchdev mode and E-Switch is enabled
we have so called VF representors in the system. All representors
belonging to the same E-Switch are created on the basis of the
single PCI function and with current implementation each representor
has its own dedicated Infiniband device and operates within its
own Infiniband context. It is proposed to provide representors
as ports of the single Infiniband device and operate on the
shared Infiniband context saving various resources. This patch
introduces appropriate structures.
Also the functions to allocate and free shared IB context for
multiport are added. The IB device context, Protection Domain,
device attributes, Infiniband names are going to be relocated
to the shared structure from the device private one.
mlx5_dev_spawn() is updated to support shared context.
net/mlx5: support multiport IB device during probing
mlx5_pci_probe() routine is refactored to probe the ports
of found Infiniband devices. All active ports (with attached
network interface), belonging to the same Infiniband device
will use the single shared Infiniband context of that device.
net/mlx5: add getting IB ports number for multiport IB
There is the routine mlx5_nl_portnum() added to get
the number of ports of multiport Infiniband device.
It is assumed the Uplink/VF representors are attached
on these ports.
net/mlx5: modify get ifindex routine for multiport IB
There is the routine mlx5_nl_ifindex() returning the
network interface index associated with Infiniband device.
We are going to support multiport IB devices, now function
takes the IB port as argument and returns ifindex associated
with tuple <IB device, IB port>
net/mlx5: add representor recognition on Linux 5.x
The master device and VF representors were distinguished by
presence of port name, master device did not have one. The new Linux
kernels starting from 5.0 provide the port name for master device
and the implemented representor recognizing method does not work.
The new recognizing method is based on querying the VF number,
has been created on the base of the device.
The IFLA_NUM_VF attribute is returned by kernel if IFLA_EXT_MASK
attribute is specified in the Netlink request message.
Also the presence check of device symlink in device sysfs folder
is added to distinguish representors with sysfs based method.
Tiwei Bie [Mon, 25 Mar 2019 05:27:16 +0000 (13:27 +0800)]
net/virtio: define avail and used flags as constants
We are consistently passing 1 as the argument in the data path,
so there is no need to define avail/used flags as function-like
macros anymore. This patch changes the avail and used flags to
constants. And a frequently used combination is also introduced.
Tiwei Bie [Mon, 25 Mar 2019 04:12:15 +0000 (12:12 +0800)]
net/virtio-user: fix multi-process support
This patch fixes the multi-process support for virtio-user.
Currently virtio-user just provides some limited secondary
process supports. Only some basic operations can be done in
secondary process on virtio-user port, e.g. getting port stats.
Actions which will trigger the communication with vhost backend
can't be done in secondary process for now, as the fds are
not synced between processes. The processing of server mode
devargs is also moved into virtio_user_dev_init().
Fixes: cdb068f031c6 ("bus/vdev: scan by multi-process channel") Fixes: ee27edbe0c10 ("drivers/net: share vdev data to secondary process") Cc: stable@dpdk.org Signed-off-by: Tiwei Bie <tiwei.bie@intel.com> Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Raslan Darawsheh [Mon, 18 Mar 2019 16:05:27 +0000 (16:05 +0000)]
net/failsafe: replace sub-device pointer with port id
In multiprocess context, the pointer to sub-device is shared between
processes. Previously, it was a pointer to per process eth_dev so
it's needed to replace this dependency.
Raslan Darawsheh [Mon, 18 Mar 2019 16:05:26 +0000 (16:05 +0000)]
net/failsafe: change back-reference from sub-device
In multiprocess context, the sub-device structure is shared
between processes. The reference to the failsafe device was
a per process pointer. It's changed to port id which is the
same for all processes.
Raslan Darawsheh [Mon, 18 Mar 2019 16:05:25 +0000 (16:05 +0000)]
net/failsafe: replace local device with shared data
In multiprocess context, the private structure is shared between
processes. The back reference from private to generic data was using
a pointer to a per process eth_dev. It's now changed to a reference of
the shared data.
Ali Alnubani [Thu, 21 Mar 2019 09:07:26 +0000 (09:07 +0000)]
net/mlx5: add missing return value check
This patch fixes the build failure with message:
drivers/net/mlx5/mlx5_ethdev.c: In function ‘mlx5_sysfs_switch_info’:
drivers/net/mlx5/mlx5_ethdev.c:1381:3:
error: ignoring return value of ‘fscanf’, declared with attribute
warn_unused_result [-Werror=unused-result]
fscanf(file, "%s", port_name);
^
Which reproduces on Ubuntu 16.04 LTS with
gcc (Ubuntu 5.4.0-6ubuntu1~16.04.10) 5.4.0 20160609.
Fixes: b2f3a3810125 ("net/mlx5: support new representor naming format") Signed-off-by: Ali Alnubani <alialnu@mellanox.com> Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com> Acked-by: Dekel Peled <dekelp@mellanox.com>
The bnxt driver is not correctly setting the receive VLAN offload
flags. When VLAN is offloaded the driver must set the
PKT_RX_VLAN_STRIPPED flag.
Actually, several drivers have the same bug, only most of the
Intel drivers look right. Any driver that sets vlan_tci is probably
stripping the tag, and should be setting RX_VLAN_STRIPPED.
To quote rte_mbuf.h:
/**
* The RX packet is a 802.1q VLAN packet, and the tci has been
* saved in in mbuf->vlan_tci.
* If the flag PKT_RX_VLAN_STRIPPED is also present, the VLAN
* header has been stripped from mbuf data, else it is still
* present.
*/
Qiming Yang [Mon, 25 Mar 2019 09:01:02 +0000 (17:01 +0800)]
net/ice: enable RSS when device init
This patch enabled RSS for UPD/TCP/SCTP+IPV4/IPV6 packets.
Signed-off-by: Qiming Yang <qiming.yang@intel.com> Acked-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com> Acked-by: Qi Zhang <qi.z.zhang@intel.com>
Qiming Yang [Mon, 25 Mar 2019 09:01:01 +0000 (17:01 +0800)]
net/ice: add safe mode
If E810 download package failed, driver need to go to safe mode.
In the safe mode, some advanced features will not be supported.
Signed-off-by: Qiming Yang <qiming.yang@intel.com> Acked-by: Paul M Stillwell Jr <paul.m.stillwell.jr@intel.com> Acked-by: Qi Zhang <qi.z.zhang@intel.com>
git.droids-corp.org / dpdk.git / log