The tree dpdk-next-tm does not exist anymore.
Traffic management and metering APIs, which are part of ethdev,
can be merged in the existing tree dpdk-next-qos.
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
Add a programmer's guide section for meson ut
Signed-off-by: Hari Kumar Vemula <hari.kumarx.vemula@intel.com>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: Michael Santana <msantana@redhat.com>
Update the documentation on building and running apps on FreeBSD, taking
account of having used meson for building. We can also update the section
on the command-line parameters, rather than claiming to be a complete list
of parameters, it should describe how to get the complete list and only
cover a few important ones.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: Anatoly Burakov <anatoly.burakov@intel.com>
The kernel modules are now installed in the correct system location on
install when using meson and ninja, so update the documentation to remove
any references to the "kmod" directory. Also, make a few additional updates
to improve clarity.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Update the FreeBSD GSG doc with instructions for installing
using meson and ninja rather than make.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: Anatoly Burakov <anatoly.burakov@intel.com>
Update the FreeBSD GSG to cover installing from the pre-built package as
well as installing from a port.
Also, since the port is now based on meson, update the instructions for
compiling and running the example applications.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Update the introduction section note, to add in the fact that DPDK is now
packaged in FreeBSD, i.e. available as a pre-built binary package, as well
as being in the ports collection for manual building.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Generate a dependency file for the header files used in the API guide
so that the docs can be rebuilt if a header changes.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: Luca Boccassi <bluca@debian.org>
For the doxygen API for the examples folder, we can generate a build
dependency file when we generate the examples.dox file. This allows
correct rebuilds if the files in examples change.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: Luca Boccassi <bluca@debian.org>
Changed the rte_ring chapter in programmer's guide to reflect
the addition of rte_ring_xxx_elem APIs. References to pointers
as ring elements is changed to generic term 'objects'.
Signed-off-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
Reviewed-by: Gavin Hu <gavin.hu@arm.com>
Correct the documentation for obj_table in rte_ring_xxx_elem APIs.
Fixes: cc4b218790f6 ("ring: support configurable element size")
Signed-off-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
Reviewed-by: Gavin Hu <gavin.hu@arm.com>
Added introduction of rte_ring_xxx_elem APIs to release notes.
Fixes: cc4b218790f6 ("ring: support configurable element size")
Signed-off-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
Skip master lcore when assigning cores to rx_queues as it is only used
to print stats.
Fixes: 4ff457986f76 ("examples/l2fwd-event: add default poll mode routines")
Cc: stable@dpdk.org
Signed-off-by: Pavan Nikhilesh <pbhagavatula@marvell.com>
Service core mask should be a hexadecimal value rather than a range of
lcores.
Bugzilla ID: 392
Fixes: 55499896d91a ("doc: add event mode to l3fwd guide")
Signed-off-by: Pavan Nikhilesh <pbhagavatula@marvell.com>
This patch sets the number of atomic ordered sequences
supported by the driver.
Fixes: dbf63bd43afa ("event/dpaa2: support ordered queue")
Cc: stable@dpdk.org
Signed-off-by: Nipun Gupta <nipun.gupta@nxp.com>
Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com>
Currently when application requests for RTE_FC_NONE earlier
than PMD's internal fc mode update, flow control set routine
is returning without updating the flow control state.
Patch updates the PMD's internal fc mode details during
dev_configure to ensure any flow control set requests
issued later are handled properly.
Fixes: 609945f1ce90 ("net/octeontx2: support flow control")
Cc: stable@dpdk.org
Signed-off-by: Vamsi Attunuru <vattunuru@marvell.com>
Acked-by: Jerin Jacob <jerinj@marvell.com>
Added support for API rte_eth_tx_done_cleanup.
Added support setting VF MAC address by PF driver.
Added support for setting the link to specific speed.
Fixes: 26e9c3e20180 ("net/ixgbe: cleanup Tx buffers")
Fixes: 3c4270187518 ("net/ixgbe: support VF MAC address add/remove")
Fixes: 34e736096068 ("net/ixgbe: set fixed flag for exact link speed")
Signed-off-by: Jiaqi Min <jiaqix.min@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>
Added support for API rte_eth_tx_done_cleanup.
Fixes: ab7cfe1fe3d7 ("net/ice: cleanup Tx buffers")
Signed-off-by: Jiaqi Min <jiaqix.min@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>
Added support for API rte_eth_tx_done_cleanup.
Added support for VF multiple queues interrupt.
Added support for setting the link to specific speed.
Fixes: 6a8defc552b1 ("net/i40e: cleanup Tx buffers")
Fixes: 40163f9e1756 ("net/i40e: fix multi-queue Rx interrupt for VF")
Fixes: dc73db27bb61 ("net/i40e: set fixed flag for exact link speed")
Signed-off-by: Jiaqi Min <jiaqix.min@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>
Coverity complains of unchecked return value warning of
i40e_xmit_cleanup, while this cleanup is opportunistic and will not
cause problems if it fails. So instead of checking the return value of
i40e_xmit_cleanup and return in case of cleanup failure, we directly
cast it to void function to make the Coverity happy.
Coverity issue: 353617
Fixes: 4861cde46116 ("i40e: new poll mode driver")
Cc: stable@dpdk.org
Signed-off-by: Beilei Xing <beilei.xing@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>
Coverity complains of unchecked return value warning of
ice_xmit_cleanup, while this cleanup is opportunistic and will not cause
problems if it fails. So instead of checking the return value of
ice_xmit_cleanup and return in case of cleanup failure, we directly cast
it to void function to make the Coverity happy.
Coverity issue: 353623
Fixes: 17c7d0f9d6a4 ("net/ice: support basic Rx/Tx")
Cc: stable@dpdk.org
Signed-off-by: Sunil Pai G <sunil.pai.g@intel.com>
Acked-by: Xiaolong Ye <xiaolong.ye@intel.com>
Currently, Rx offload capabilities and max_rx_pkt_len in the struct
variable named rte_port are not updated after setting mtu successfully
in port_mtu_set function by 'port config mtu <port_id> <value>' command.
This may lead to reconfig mtu to the initial value in the driver when
recalling rte_eth_dev_configure API interface.
This patch updates Rx offload capabilities and max_rx_pkt_len after
setting mtu successfully when configuring mtu.
Fixes: ae03d0d18adf ("app/testpmd: command to configure MTU")
Cc: stable@dpdk.org
Signed-off-by: Huisong Li <lihuisong@huawei.com>
Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
There is a possible race condition in the hotplug path
in rmv_port_callback(). If a port is created between
close_port(port_id) and detach_port_device(port_id),
then the port_id will have been reallocated to a different
device which will be wrongly detached.
Since a check was added in detach_port_device() for
manual detach case, the hotplug path was even more broken.
It became impossible to run because the new check prevented
to run detach_port_device() after the port is closed.
The solution for both issues is to not rely on the port_id
for detaching the rte_device.
The function detach_port_device() is split to allow calling
detach_device() directly with the rte_device pointer, saved
before closing the port.
Fixes: 43d0e304980a ("app/testpmd: fix invalid port detaching")
Cc: stable@dpdk.org
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
There is a function detach_port_device() which takes a port_id,
and a function detach_device() which takes a devargs string.
In order to add a third function accepting a rte_device pointer,
the function detach_device() is renamed into detach_devargs().
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
User may try to run "port detach <port_id>"
for an already detached device.
It has been decided to protect from such usage in testpmd,
so a check was added to detach_port_device() in DPDK 19.11.
This check might be removed to allow hotplug path detaching
the device of a closed port.
Whatever will be decided in future, this check is also added
before the call to detach_port_device().
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
Currently, the log address translation only happens in the vhost-user's
translate_ring_addresses(). However, the IOTLB update handler is not
checking if it was mapped to re-trigger that translation.
Since the log address mapping could fail, check it on iotlb updates.
Also, check it on vring_translate() so we do not dirty pages if the
logging address is not yet ready.
Additionally, properly protect the accesses to the iotlb structures.
Fixes: fbda9f145927 ("vhost: translate incoming log address to GPA")
Cc: stable@dpdk.org
Signed-off-by: Adrian Moreno <amorenoz@redhat.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
The frontend may not send the get_inflight_fd and
set_inflight_fd although we negotiate the protocol
feature. When we meet this situation just return OK.
Fixes: ad0a4ae491fe ("vhost: checkout resubmit inflight information")
Cc: stable@dpdk.org
Signed-off-by: Jin Yu <jin.yu@intel.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Add release notes for the notification data feature of virtio-PMD.
Fixes: 7e72f3ec1a8a ("net/virtio: support packed ring notification data")
Signed-off-by: Cheng Jiang <cheng1.jiang@intel.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Fix the time of check time of use warning in example code.
Ignore the errno of unlink failure. There are two situations.
The first one is that file doesn't exist the unlink fails and
it's ok to ignore. The second one is that unlink fails to remove
file but the next bind() would fail too.
Coverity issue: 350589, 158663
Fixes: c19beb3f38cd ("examples/vhost_blk: introduce vhost storage sample")
Cc: stable@dpdk.org
Signed-off-by: Jin Yu <jin.yu@intel.com>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
This patch removes an unused member from pmd_internal.
Signed-off-by: Itsuro Oda <oda@valinux.co.jp>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
This patch adds lacking member setting and makes secondary
probe complete.
Fixes: 4852aa8f6e21 ("drivers/net: enable hotplug on secondary process")
Cc: stable@dpdk.org
Signed-off-by: Itsuro Oda <oda@valinux.co.jp>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Vhost driver setup is delayed at eth_dev configuration
in order to be able to set it from a secondary process.
Fixes: 4852aa8f6e21 ("drivers/net: enable hotplug on secondary process")
Cc: stable@dpdk.org
Signed-off-by: Itsuro Oda <oda@valinux.co.jp>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
This patch allocates iface_name of pmd_internal from heap
in order to be able to refer from secondary processes.
Fixes: 4852aa8f6e21 ("drivers/net: enable hotplug on secondary process")
Cc: stable@dpdk.org
Signed-off-by: Itsuro Oda <oda@valinux.co.jp>
Reviewed-by: Maxime Coquelin <maxime.coquelin@redhat.com>
This patch adds a check to ensure the read size of
the Vhost-user message header is not smaller than
the expected size.
In case of unexpected read size, report an error
and close file descriptors passed with the message,
if any.
Fixes: 8f972312b8f4 ("vhost: support vhost-user")
Cc: stable@dpdk.org
Reported-by: Ilja Van Sprundel <ivansprundel@ioactive.com>
Signed-off-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Reviewed-by: Tiwei Bie <tiwei.bie@intel.com>
If ibverbs_link is dlopen, the PMD and application should not
be linked with ibverbs, but the glue library is.
Unfortunately the ibverbs dependency was exported in the
variable ext_deps, so there were overlinking.
It is fixed by not exporting the dependency in ext_deps,
and recreating a limited dependency object for cflags only.
Fixes: 1dd7c7e38c19 ("net/mlx4: support meson build")
Fixes: 96d7c62a70c7 ("net/mlx5: support meson build")
Cc: stable@dpdk.org
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
The name of the variable pmd_dlopen is confusing because
it can be understood as true if the PMD is dlopen,
whereas it means the ibverbs glue layer is a dlopen library.
That's why it is renamed dlopen_ibverbs.
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
If ibverbs_link is static and the application choose to link DPDK
as static libraries, both PMD and ibverbs libraries must be linked
as static libraries. And the dependencies of ibverbs (netlink) must
still be linked as shared libraries.
Unfortunately, meson forget about the static requirement for ibverbs
when generating the .pc file.
As a result, libibverbs, libmlx4, libmlx5 are listed in Requires.private
section (allowing to be linked as shared libraries) and libnl is missing.
A fix is in progress for meson, but anyway we will have to live without
such a fix until a better version of meson is widely available:
https://github.com/mesonbuild/meson/pull/6393
In order to avoid meson suggesting shared libraries in the section
Requires.private of the .pc file, the dependency object is recreated
with declare_dependency():
- cflags are extracted the libibverbs.pc
- ldflags, from libibverbs.pc, are processed to force
static flavor of ibverbs libraries, thanks to this syntax:
-l:libfoo.a
Fixes: 6affeabaf321 ("net/mlx: add static ibverbs linkage with meson")
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
The shell script options-ibverbs-static.sh was used with make
in forcing static linkage of ibverbs libraries.
When choosing to link with a static dependency in meson,
the generated .pc file will not force such static linkage.
The solution will rely on using this script in meson.
If linking with libraries installed in a non-standard path,
an option -L is provided via EXTRA_LDFLAGS in case of using make.
With meson, tuning PKG_CONFIG_PATH for pkg-config should be enough.
When statically linking an application, the .pc file must save the
-L path so the application link will work without any extra option.
That's why --libs-only-l is replaced with --libs which includes -L.
Options which are neither -l or -L are filtered out because not needed
and can cause compilation issues with the legacy system using make.
The other change in this script is to drop the first occurrences of the
main library file (libiverbs.a). Only the last occurrence is kept.
It fixes some undefined references when linking a static application
using libdpdk.pc.
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
The libibverbs (and libmlx4/5) can be statically embedded
in the shared PMD library, or in the application with the static PMD.
It was supported with make build system in
commit 2c0dd7b69fb0 ("config: add static linkage of mlx dependency").
The same feature is enabled with meson when using pkg-config
(i.e. only if the call to dependency() is successful).
The fallback method for searching library with cc.find_library()
is not supported because the dependencies of the found library
would not be linked (no such info in .a file unlike .so).
The main difference, in meson build system, is the generated .pc file
giving arguments to link DPDK with the application.
Unfortunately the .pc file will not keep memory of the static linkage
option for libibverbs.
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
The variable build is already initialized as true in
drivers/meson.build. Duplicate initializations can be removed from mlx.
When the variable build is set to false, it is easier to call
subdir_done() than branch the rest of the code on build condition.
Suggested-by: Bruce Richardson <bruce.richardson@intel.com>
Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Acked-by: Rosen Xu <rosen.xu@intel.com>
The glue shared library name was created by the common class FMT name,
driver_name_fmt = 'rte_common_@0@', which is not correlated with
LIB_GLUE_BASE = 'librte_pmd_mlx5_glue.so'.
This causes the dynamic linkage of the glue library to fail what
remained mlx5 drivers outside the PCI driver list.
The command:
MLX5_GLUE_PATH=$(pwd)/build-meson/drivers/common/mlx5
build-meson/app/dpdk-testpmd -n 4 -w 0000:00:05.0 -w 0000:00:06.0 -- -i
The log:
common_mlx5: Cannot load glue library:
/usr/local/lib64/dpdk/pmds-20.0.1-glue/librte_pmd_mlx5_glue.so.20.02.0:
cannot open shared object file: No such file or directory
Adjust the shared library name to the LIB_GLUE_BASE definition.
Fixes: 7b4f1e6bd367 ("common/mlx5: introduce common library")
Signed-off-by: Matan Azrad <matan@mellanox.com>
The maximum MTU for af_xdp zero copy is equal to the page size less the
frame overhead introduced by AF_XDP (XDP HR = 256) and DPDK (frame
headroom = 320). The patch updates this value to reflect this.
This change also makes it possible to remove unneeded constants for both
zero-copy and copy mode.
Fixes: d8a210774e1d ("net/af_xdp: support unaligned umem chunks")
Cc: stable@dpdk.org
Signed-off-by: Ciara Loftus <ciara.loftus@intel.com>
Reviewed-by: Xiaolong Ye <xiaolong.ye@intel.com>
The fill queue addresses should start at the beginning of the mempool
object instead of the beginning of the mbuf. This is because the umem
frame headroom includes the mp hdrobj size. Starting at this point
ensures AF_XDP doesn't write past the available room in the frame, in
the case of larger packets which are close to the size of the mbuf.
Fixes: d8a210774e1d ("net/af_xdp: support unaligned umem chunks")
Cc: stable@dpdk.org
Signed-off-by: Ciara Loftus <ciara.loftus@intel.com>
Reviewed-by: Xiaolong Ye <xiaolong.ye@intel.com>
The previous frame size calculation incorrectly used
mb_pool->private_data_size and didn't include mb_pool->header_size.
Instead of performing a manual calculation, use the
rte_mempool_calc_obj_size API to determine the frame size.
The previous frame headroom calculation also incorrectly used
mb_pool->private_data_size and didn't include mb_pool->header_size or
the mbuf priv size. Fix this.
Fixes: d8a210774e1d ("net/af_xdp: support unaligned umem chunks")
Cc: stable@dpdk.org
Signed-off-by: Ciara Loftus <ciara.loftus@intel.com>
Reviewed-by: Xiaolong Ye <xiaolong.ye@intel.com>
In current version, we are setting the ports
using portmask. With portmask, we can use only
up to 64 ports. This portlist option enables the user
to use more than 64 ports.
Now we can specify the ports in 2 different ways
- Using portmask (-p [0x]nnn): mask must be in hex format
- Using portlist in the following format
--portlist <p1>[-p2][,p3[-p4],...]
--portmask 0x2 is same as --portlist 1
--portmask 0x3 is same as --portlist 0-1
Signed-off-by: Hariprasad Govindharajan <hariprasad.govindharajan@intel.com>
Acked-by: Anatoly Burakov <anatoly.burakov@intel.com>
Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
Reviewed-by: Herakliusz Lipiec <herakliusz.lipiec@intel.com>
The meter flows are split into three subflows each, the prefix subflow
with meter action color the packet, the meter subflow filters out the
colored packets, the suffix subflow applies all the remaining actions
to the passed packets. The tag header modify action is added to the
prefix subflow to make the suffix subflow to match the packets from the
prefix subflow.
Currently, the tag header modify action is added at the beginning in the
prefix subflow even before decap action. The header modify action does
not make sense to the later decap action, so the flow create will be
validated as incorrect flow rule and failed.
Move the tag header modify action just before meter action in the prefix
subflow to make the flow with decap action to do the decap first, then
do the tag and meter to fix that issue.
Fixes: 9ea9b049a960 ("net/mlx5: split meter flow")
Cc: stable@dpdk.org
Reported-by: Tonghao Zhang <xiangxia.m.yue@gmail.com>
Signed-off-by: Suanming Mou <suanmingm@mellanox.com>
Acked-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com>
Acked-by: Matan Azrad <matan@mellanox.com>
Currently MLX5 PMD can't match on untagged packets specifically.
Tagged traffic still hits the flows intended for untagged packets.
If the flow has ETH, it will catch all matching packets, tagged
and untagged.
The solution is to use cvlan_tag bit.
If mask=1 and value=0 it matches on untagged traffic.
If mask=1 and value=1 it matches on tagged traffic.
This is the kernel implementation.
This patch updated MLX5 PMD to set cvlan_tag mask and value according
to flow rule contents.
This update is relevant when using DV flow engine (dv_flow_en=1).
See example at https://doc.dpdk.org/guides/nics/mlx5.html#limitations.
Fixes: fc2c498ccb94 ("net/mlx5: add Direct Verbs translate items")
Cc: stable@dpdk.org
Signed-off-by: Dekel Peled <dekelp@mellanox.com>
Acked-by: Matan Azrad <matan@mellanox.com>
