Stefan Hajnoczi [Wed, 31 Jan 2018 17:46:50 +0000 (17:46 +0000)]
vhost: add flag for built-in virtio driver
The librte_vhost API is used in two ways:
1. As a vhost net device backend via rte_vhost_enqueue/dequeue_burst().
2. As a library for implementing vhost device backends.
There is no distinction between the two at the API level or in the
librte_vhost implementation. For example, device state is kept in
"struct virtio_net" regardless of whether this is actually a net device
backend or whether the built-in virtio_net.c driver is in use.
The virtio_net.c driver should be a librte_vhost API client just like
the vhost-scsi code and have no special access to vhost.h internals.
Unfortunately, fixing this requires significant librte_vhost API
changes.
This patch takes a different approach: keep the librte_vhost API
unchanged but track whether the built-in virtio_net.c driver is in use.
See the next patch for a bug fix that requires knowledge of whether
virtio_net.c is in use.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com> Acked-by: Yuanhan Liu <yliu@fridaylinux.org>
Shahaf Shuler [Thu, 1 Feb 2018 18:53:53 +0000 (20:53 +0200)]
net/mlx5: fix CRC strip capability query
IBV_WQ_FLAGS_SCATTER_FCS is a WQ flag to be used to configure the CRC
strip on a queue upon creation.
Using IBV_RAW_PACKET_CAP_SCATTER_FCS instead to query to capability.
Even though this is RAW_QP capability, it is being used by rdma-core to
indicate for both RAW_QP and WQ.
Fixes: 43e9d9794cde ("net/mlx5: support upstream rdma-core") Cc: stable@dpdk.org Reported-by: Alex Rosenbaum <alexr@mellanox.com> Signed-off-by: Shahaf Shuler <shahafs@mellanox.com> Acked-by: Yongseok Koh <yskoh@mellanox.com>
Nélio Laranjeiro [Tue, 30 Jan 2018 13:36:52 +0000 (14:36 +0100)]
net/mlx5: fix port stop by verify flows are still present
priv_flow_stop() may be called several times, in such situation flows are
already removed from the NIC and thus all associated objects are no present
in the flow object (ibv_flow, indirection tables, ....).
Adrien Mazarguil [Wed, 31 Jan 2018 15:33:06 +0000 (16:33 +0100)]
net/mlx4: fix drop flow resources leak
Resources allocated for drop flow rules are not freed properly. This causes
a memory leak and triggers an assertion failure on a reference counter when
compiled in debug mode.
This issue can be reproduced with testpmd by entering the following
commands:
flow create 0 ingress pattern eth / end actions drop / end
port start all
port stop all
port start all
port stop all
quit
The reason is additional references are taken when re-enabling existing
flow rules, a common occurrence when rehashing configuration.
Fixes: d3a7e09234e4 ("net/mlx4: allocate drop flow resources on demand") Cc: stable@dpdk.org Reported-by: Moti Haimovsky <motih@mellanox.com> Signed-off-by: Adrien Mazarguil <adrien.mazarguil@6wind.com>
Beilei Xing [Fri, 2 Feb 2018 12:05:52 +0000 (20:05 +0800)]
net/i40e: fix interrupt conflict with multi-driver
There's interrupt conflict when using DPDK and Linux i40e
on different ports of the same Ethernet controller, this
patch fixes it by switching from IntN to Int0 if multiple
drivers are used.
Beilei Xing [Fri, 2 Feb 2018 12:05:51 +0000 (20:05 +0800)]
net/i40e: fix multiple driver support
This patch provides the option to disable writing some global registers
in PMD, in order to avoid affecting other drivers, when multiple drivers
run on the same NIC and control different physical ports. Because there
are few global resources shared among different physical ports.
Fixes: ec246eeb5da1 ("i40e: use default filter input set on init") Fixes: 98f055707685 ("i40e: configure input fields for RSS or flow director") Fixes: f05ec7d77e41 ("i40e: initialize flow director flexible payload setting") Fixes: e536c2e32883 ("net/i40e: fix parsing QinQ packets type") Fixes: 19b16e2f6442 ("ethdev: add vlan type when setting ether type") Cc: stable@dpdk.org Signed-off-by: Beilei Xing <beilei.xing@intel.com> Acked-by: Jingjing Wu <jingjing.wu@intel.com>
Xiaoyun Li [Fri, 2 Feb 2018 05:45:16 +0000 (13:45 +0800)]
net/i40e: fix CRC strip enabling
New testpmd will get CRC strip offload from rx_offload_capa. I40evf
cannot disable CRC strip. And in fact, it is enabled by PF. This
patch solves the issue by adding CRC strip flag into rx_offload_capa
in i40e and i40evf.
Fixes: 8b9bd0efe0b6 ("app/testpmd: disable Rx VLAN offloads by default") Signed-off-by: Xiaoyun Li <xiaoyun.li@intel.com> Acked-by: Beilei Xing <beilei.xing@intel.com>
Ivan Malov [Sat, 3 Feb 2018 13:46:33 +0000 (13:46 +0000)]
net/sfc: rename version map file
The version map filename does not comply with the format
used by meson build rules for drivers (i.e. on the upper
level) and needs to be revisited. This patch removes efx
postfix from the driver title in the filename.
Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Acked-by: Bruce Richardson <bruce.richardson@intel.com>
The power management and KNI libraries are not compiled on a FreeBSD
platform, which means that the tests can't run. Add in stub code for
these cases, allowing the tests to still be compiled, but to report
as skipped in those cases.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: David Hunt <david.hunt@intel.com>
build: fix dependency on execinfo for BSD meson builds
The binaries and apps in DPDK all need to be linked against the
execinfo library on FreeBSD so add this as a dependency in cases
where it is found. It's available by default on BSD, but not
at all on Linux
Fixes: 16ade738fd0d ("app/testpmd: build with meson") Fixes: 89f0711f9ddf ("examples: build some samples with meson") Fixes: b5dc795a8a55 ("test: build app with meson as dpdk-test") Fixes: 2ff67267b049 ("app/eventdev: build with meson") Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
The kernel module source file directory passed via VPATH was wrong,
which caused the source files to be not found via make. Rather than
explicitly passing VPATH, make use of the fact that the full path
to the source files is passed by meson, so split that into directory
part - to be used as VPATH - and file part - to be used as the source
filename.
Fixes: 610beca42ea4 ("build: remove library special cases") Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Bruce Richardson [Wed, 31 Jan 2018 17:09:05 +0000 (17:09 +0000)]
net/i40e: fix dependencies for AVX2 build with meson
The AVX2 code path includes files from the ethdev, hash and kvargs libs.
These are not listed as dependencies in the case where AVX2 is not in
the default instruction set for the build e.g. machine=nehalem. This
leads to compiler errors as the header files needed cannot be found.
Fixes: e940646b20fa ("drivers/net: build Intel NIC PMDs with meson") Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Anatoly Burakov [Wed, 24 Jan 2018 10:26:12 +0000 (10:26 +0000)]
usertools/devbind: fix kernel module reporting
lspci reports kernel modules in "Module" string, but devbind
expects it to be "Module_str". Fix it up similar to how we fix
up "Driver" to be "Driver_str".
Currently, memzone autotest expects there to be no memzones
present by the time the test is run. Some hardware drivers
will allocate memzones for internal use during initialization,
resulting in tests failing due to unexpected memzones being
allocated before the test was run.
Fix this by making sure all memzones allocated by this test
have a common prefix, and making callback increment a counter
on encountering memzones with this prefix. Also, separately
increment another counter that will count total number of
memzones left after test, and compares it to previously stored
number of memzones, to ensure that we didn't accidentally
allocated/freed any memzones we weren't supposed to. This
also doubles as a test for correct operation of memzone_walk().
Fixes: 71330483a193 ("test/memzone: fix memory leak") Cc: stable@dpdk.org Signed-off-by: Phil Yang <phil.yang@arm.com> Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com> Reviewed-by: Radoslaw Biernacki <radoslaw.biernacki@linaro.org>
Anatoly Burakov [Wed, 31 Jan 2018 14:17:32 +0000 (14:17 +0000)]
test: clean up memory for function reentrancy test
Function reentrancy test limits maximum number of iterations based
on the number of memzones and cores, however it doesn't free the
memzones after the fact, so on a machine with big amount of cores
the tests will fail due to running out of memzones.
Fix this by introducing cleanup functions for ring and mempool
reentrancy tests.
Coverity flags an issue where the resources used by the FILE object for
the temporary input file are leaked. This is a very minor issue, but is
easily fixed, while also avoiding later problems where we try to close
an invalid file descriptor in the failure case.
The fix is to use "dup()" to get a new file descriptor number rather than
using the value directly from fileno. This allows us to close the file
opened with tmpfile() within in scope block, while allowing the duplicate
to pass to the outer block and be closed when the function terminates.
As a side-effect I/O in the function is therefore changed from using stdio
fread/fwrite to read/write system calls.
Coverity issue: 260399 Fixes: 0d68533617e3 ("pmdinfogen: allow using stdin and stdout") Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: Neil Horman <nhorman@tuxdriver.com>
Nipun Gupta [Fri, 2 Feb 2018 08:03:01 +0000 (13:33 +0530)]
mbuf: fix logic of user mempool ops API
The existing rte_eal_mbuf_default mempool ops can return the compile time
default ops name if the user has not provided command line inputs for
mempool ops name. It will break the logic of best mempool ops as it will
never return platform hw mempool ops.
This patch introduces a new API to just return the user mempool ops only.
Fixes: 8b0f7f434132 ("mbuf: maintain user and compile time mempool ops name") Signed-off-by: Nipun Gupta <nipun.gupta@nxp.com> Acked-by: Olivier Matz <olivier.matz@6wind.com>
There is not specified dependency between rte_mempool_populate_default()
and rte_mempool_populate_iova(). So, the second should not rely on the
fact that the first adds capability flags to the mempool flags.
Tonghao Zhang [Tue, 23 Jan 2018 16:17:34 +0000 (08:17 -0800)]
igb_uio: print IRQ as decimal number
The kernel uses the '%d' or '%ld' to print irq num.
But igb_uio may use the '%lx', then the log may confuse
the user what irq num has been used. The log is show as
below.
igb_uio 0000:00:03.0: irq 24 for MSI/MSI-X
igb_uio 0000:00:03.0: uio device registered with irq 18
kernel version: 3.10.0-514.16.1.el7
For avoiding to be confused, change the igb_uio irq
print type.
Marko Kovacevic [Mon, 5 Feb 2018 15:45:31 +0000 (15:45 +0000)]
mk: fix external build
Code commit for 'make -f' support, breaks the build in cases where
entries in $(MAKEFILE_LIST) are absolute paths. This commit uses
notdir and firstword to ensure that only the local filename is used.
Fixes: 3a5c339d51a4 ("mk: support renamed Makefile in external project") Cc: stable@dpdk.org Signed-off-by: Marko Kovacevic <marko.kovacevic@intel.com>
Vipin Varghese [Tue, 30 Jan 2018 12:17:02 +0000 (17:47 +0530)]
doc: add procinfo restriction with shared library
added note category to procinfo guide, to intimate users in shared
library mode. Procinfo requires same NIC libraries to probe the NIC
devices as Primary Application.
Signed-off-by: Vipin Varghese <vipin.varghese@intel.com> Acked-by: Marko Kovacevic <marko.kovacevic@intel.com>
Junjie Chen [Wed, 29 Nov 2017 17:17:13 +0000 (12:17 -0500)]
doc: add a restriction to multi-process support
This patch add a restriction to multi-process support: secondary
processes should only run alongside primary process with same DPDK
version, so that secondary processes can use the same hugepage mmap
layout as primary process.
Signed-off-by: Junjie Chen <junjie.j.chen@intel.com> Acked-by: John McNamara <john.mcnamara@intel.com>
Bruce Richardson [Tue, 23 Jan 2018 14:11:33 +0000 (14:11 +0000)]
event/opdl: fix license header and SPDX tags
This patch ensures that the OPDL files all contain correct SPDX tags.
The following changes were made to achieve this:
* replace license text with SPDX tag
* correct occurences where SPDX tag was not on first line of file
* ensure license years were correct
Fixes: 4236ce9bf5bf ("event/opdl: add OPDL ring infrastructure library") Fixes: 3c7f3dcfb099 ("event/opdl: add PMD main body and helper function") Fixes: d548ef513cd7 ("event/opdl: add unit tests") Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: Liang Ma <liang.j.ma@intel.com>
Radu Nicolau [Wed, 24 Jan 2018 13:16:46 +0000 (13:16 +0000)]
examples/bond: check mbuf allocation
Coverity issue: 257008 Fixes: cc7e8ae84faa ("examples/bond: add example application for link bonding mode 6") Cc: stable@dpdk.org Signed-off-by: Radu Nicolau <radu.nicolau@intel.com>
David Coyle [Tue, 23 Jan 2018 14:14:15 +0000 (14:14 +0000)]
examples/vm_power_manager: fix set VF MAC address
Current code only sets mac address of first VF. Fix code so that it
continues through the loop and sets the mac address of each VF.
Fixes: c9a4779135c9 ("examples/vm_power_mgr: set MAC address of VF") Cc: stable@dpdk.org Signed-off-by: David Coyle <david.coyle@intel.com> Acked-by: David Hunt <david.hunt@intel.com>
Hemant Agrawal [Thu, 18 Jan 2018 06:12:58 +0000 (11:42 +0530)]
kni: support MAC address change
This patch adds following:
1. Option to configure the mac address during create. Generate random
address only if the user has not provided any valid address.
2. Inform usespace, if mac address is being changed in linux.
3. Implement default handling of mac address change in the corresponding
ethernet device.
Pavan Nikhilesh [Tue, 30 Jan 2018 18:51:54 +0000 (00:21 +0530)]
lpm: fix allocation of an existing object
Fix rte_lpm_create_*() functions to return NULL and set rte_errno to
EEXIST when lpm object name already exists.
This is the behavior described in the API documentation in the header
file.
Yong Wang [Thu, 25 Jan 2018 06:53:04 +0000 (01:53 -0500)]
crypto/qat: fix allocation check and leak
There are several func calls to rte_zmalloc() which don't do null
point check on the return value. And before return, the memory is not
freed. Fix it by adding null point check and rte_free().
Vipin Varghese [Fri, 26 Jan 2018 22:27:26 +0000 (03:57 +0530)]
net/tap: remove speed argument
TAP is a virtual device created on Kernel. The speed of interface is
set by Kernel to a fixed static value. But this does not prevent using
RX or TX to rate limit. Hence removing the option from user arguments.
Radu Nicolau [Thu, 25 Jan 2018 16:26:26 +0000 (16:26 +0000)]
examples/bond: fix vdev name
Fixes: 73db5bad ("net: align ethdev and eal driver names") Cc: stable@dpdk.org Signed-off-by: Radu Nicolau <radu.nicolau@intel.com> Acked-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>
Ophir Munk [Tue, 30 Jan 2018 16:00:29 +0000 (16:00 +0000)]
net/tap: fix eBPF RSS map key handling
This commit addresses a case of RSS and non RSS flows mixture.
In the corrupted code, if a non-RSS flow is destroyed, it has an eBPF map
index 0 (initialized upon flow creation). This might unintentionally remove
RSS entry 0 from eBPF map.
To fix this issue, add an offset to the real index during a KEY_CMD_GET
operation, and subtract this offset during a KEY_CMD_RELEASE operation, in
order to restore the real index.
Thus, if a non RSS flow is falsely trying to release map entry 0 - The
offset subtraction will calculate the real map index as an
out-of-range value, and the release operation will be silently ignored.
Ophir Munk [Tue, 30 Jan 2018 16:00:28 +0000 (16:00 +0000)]
net/tap: fix eBPF file descriptors leakage
When a user creates an RSS rule, the tap PMD dynamically allocates
a 'flow' data structure, and uploads BPF programs (represented by file
descriptors) to the kernel.
The kernel might reject the rule (due to filters overlap, for example)
in which case, flow memory should be freed and BPF file descriptors
should be closed.
In the corrupted code there were scenarios where BPF file descriptors
were not closed.
The fix is to add a new function - tap_flow_free(), which will make sure
to always close BPF file descriptors before freeing the flow allocated
memory.
Adrien Mazarguil [Tue, 30 Jan 2018 15:34:58 +0000 (16:34 +0100)]
net/mlx5: spawn rdma-core dependency plug-in
When mlx5 is not compiled directly as an independent shared object (e.g.
CONFIG_RTE_BUILD_SHARED_LIB not enabled for performance reasons), DPDK
applications inherit its dependencies on libibverbs and libmlx5 through
rte.app.mk.
This is an issue both when DPDK is delivered as a binary package (Linux
distributions) and for end users because rdma-core then propagates as a
mandatory dependency for everything.
Application writers relying on binary DPDK packages are not necessarily
aware of this fact and may end up delivering packages with broken
dependencies.
This patch therefore introduces an intermediate internal plug-in
hard-linked with rdma-core (to preserve symbol versioning) loaded by the
PMD through dlopen(), so that a missing rdma-core does not cause unresolved
symbols, allowing applications to start normally.
Adrien Mazarguil [Tue, 30 Jan 2018 15:34:54 +0000 (16:34 +0100)]
net/mlx4: spawn rdma-core dependency plug-in
When mlx4 is not compiled directly as an independent shared object (e.g.
CONFIG_RTE_BUILD_SHARED_LIB not enabled for performance reasons), DPDK
applications inherit its dependencies on libibverbs and libmlx4 through
rte.app.mk.
This is an issue both when DPDK is delivered as a binary package (Linux
distributions) and for end users because rdma-core then propagates as a
mandatory dependency for everything.
Application writers relying on binary DPDK packages are not necessarily
aware of this fact and may end up delivering packages with broken
dependencies.
This patch therefore introduces an intermediate internal plug-in
hard-linked with rdma-core (to preserve symbol versioning) loaded by the
PMD through dlopen(), so that a missing rdma-core does not cause unresolved
symbols, allowing applications to start normally.
Wei Zhao [Fri, 26 Jan 2018 08:46:51 +0000 (16:46 +0800)]
net/i40e: fix flow RSS return error
This patch fixes issues reported by Coverity check. Function
parse_rss_action and i40e_config_rss_filter might return at wrong
time which will cause error for RSS configuration and parser. Hash
function variable with 32 bits width might also cause error when
it needs more than 32 bits, so change this variable to 64 bits.
This patch moves the definition of 3 variables in testpmd.h
into the respective .c file. The idea behind this move is
to allow external applications to compile against testpmd
without throwing compilation errors related to multiple
definition of variables.
Also, an extern dcb_q_mapping in testpmd.h is removed
since it appears that this variable is not defined
elsewhere in the tree.
Remy Horton [Fri, 19 Jan 2018 13:27:22 +0000 (13:27 +0000)]
app/testpmd: fix port id size
The word size of port_id is now 16 bits, but there were parsing directives
that assumed it was still of type UINT8, resulting in incorrect commandline
parse results.
Fixes: f14a210a65fe ("app: fix port id type") Cc: stable@dpdk.org Signed-off-by: Remy Horton <remy.horton@intel.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
Jiayu Hu [Thu, 25 Jan 2018 02:13:44 +0000 (10:13 +0800)]
app/testpmd: fix SW checksum calculation enabling
In current design, we can't enable SW checksum calculation
for the devices which don't have checksum offloading abilities
via the command "csum set ip|tcp|udp|sctp|outer-ip sw <port_id>".
But SW checksum calculation shouldn't depend on HW offloading
abilities. This patch is to fix this issue.
Moti Haimovsky [Wed, 31 Jan 2018 17:32:03 +0000 (19:32 +0200)]
app/testpmd: disable Rx VLAN offloads by default
Removed the hardcoded preconfigured Rx VLAN offload configuration
from testpmd and changed the Rx offload command line parameters from
disable to enable.
It has been decided by the Technical Board that testers who wish to
use these offloads will now have to explicitly write them in the
command-line when running testpmd.
The agreement is to keep two exceptions enabled by default in 18.02:
Rx CRC strip and Tx fast free.
Motivation:
Some PMDs such at the mlx4 may not implement all the offloads.
After the offload API rework assuming no offload is enabled by default,
commit ce17eddefc20 ("ethdev: introduce Rx queue offloads API")
commit cba7f53b717d ("ethdev: introduce Tx queue offloads API") trying
to enable a not supported offload is clearly an error which will cause
configuration failing.
Considering that testpmd is an application to test the PMD, it should
not fail on a configuration which was not explicitly requested.
The behavior of this test application is then turned to an opt-in
model.
Signed-off-by: Moti Haimovsky <motih@mellanox.com> Acked-by: Thomas Monjalon <thomas@monjalon.net>
Matan Azrad [Tue, 30 Jan 2018 14:13:40 +0000 (14:13 +0000)]
app/testpmd: fix invalid port detach
Using registration to all ports includes also the ports which should not
be used by the application.
It is nice to print each port event by testpmd but in case of RMV
event, testpmd tries to detach the port and this case is problematic
when the port should not be used by the application.
Check the port validation before detach API calling.
git.droids-corp.org / dpdk.git / log