Suanming Mou [Thu, 3 Dec 2020 02:18:52 +0000 (04:18 +0200)]
net/mlx5: optimize hash list entry memory
Currently, the hash list saves the hash key in the hash entry. And the
key is mostly used to get the bucket index only.
Save the entire 64 bits key to the entry will not be a good option if
the key is only used to get the bucket index. Since 64 bits costs more
memory for the entry, mostly the signature data in the key only uses
32 bits. And in the unregister function, the key in the entry causes
extra bucket index calculation.
This commit saves the bucket index to the entry instead of the hash key.
For the hash list like table, tag and mreg_copy which save the signature
data in the key, the signature data is moved to the resource data struct
itself.
Suanming Mou [Thu, 3 Dec 2020 02:18:51 +0000 (04:18 +0200)]
net/mlx5: optimize hash list synchronization
Since all the hash table operations are related with one dedicated
bucket, the hash table lock and gen_cnt can be allocated per-bucket.
Currently, the hash table uses one global lock to protect all the
buckets, that global lock avoids the buckets to be operated at one
time, it hurts the hash table performance. And the gen_cnt updated
by the entire hash table causes incorrect redundant list research.
This commit optimized the lock and gen_cnt to bucket solid allows
different bucket entries can be operated more efficiently.
Dekel Peled [Tue, 24 Nov 2020 13:45:35 +0000 (15:45 +0200)]
net/mlx5: fix shared age action validation
Previous patch added support of shared age action.
This feature is supported on group 1 and higher, and validation was
added accordingly.
On FDB table the group 0 is skipped to improve performance.
As a result the mentioned validation is not relevant for transfer rules.
This patch adds the required check to ensure proper validation.
The rdma-core library uses callbacks to allocate and free memory
from DPDK. The memory allocation callback used the complicated
and incorrect way to get the NUMA socket ID from the context.
The context was wrong that might result in wrong socket ID
and allocating memory from wrong node.
The callbacks are assigned once as Infinibande device context
is created allowing early access to shared DPDK memory for all
Verbs internal objects need that.
Adding below APIs for axgbe
- axgbe_enable_rx_vlan_stripping: to enable vlan header stripping
- axgbe_disable_rx_vlan_stripping: to disable vlan header stripping
- axgbe_enable_rx_vlan_filtering: to enable vlan filter mode
- axgbe_disable_rx_vlan_filtering: to disable vlan filter mode
- axgbe_update_vlan_hash_table: crc calculation and hash table update
based on vlan values post filter enable
- axgbe_vlan_filter_set: setting of active vlan out of max 4K values
before doing hash update of same
- axgbe_vlan_tpid_set: setting of default tpid values
- axgbe_vlan_offload_set: a top layer function to call strip/filter etc
based on mask values
Ivan Malov [Tue, 1 Dec 2020 07:30:10 +0000 (10:30 +0300)]
common/sfc_efx/base: support alternative MAE match fields
If MAE slice is configured without conntrack support, outer
rules must match on IP SRC/DST. This isn't reported clearly
by the FW because IPv4 and IPv6 have separate SRC/DST pairs.
The FW reports status ALWAYS for all these four fields, and
having an all-zeros mask for either field prevents the spec
from being certified by the existing spec validation method.
Extend the spec validation to take the "alternative" fields
into account so that legitimate specs don't get turned down.
Fixes: ed15d7f8e064 ("common/sfc_efx/base: validate and compare outer match specs") Cc: stable@dpdk.org Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru> Reviewed-by: Andrew Rybchenko <andrew.rybchenko@oktetlabs.ru> Reviewed-by: Andy Moreton <amoreton@xilinx.com>
Currently, the queue stats mapping has the following problems:
1) Many PMD drivers don't support queue stats mapping. But there is no
failure message after executing the command "set stat_qmap rx 0 2 2".
2) Once queue mapping is set, unrelated and unmapped queues are also
displayed.
3) The configuration result does not take effect or can not be queried
in real time.
4) The mapping arrays, "tx_queue_stats_mappings_array" &
"rx_queue_stats_mappings_array" are global and their sizes are based
on fixed max port and queue size assumptions.
5) These record structures, 'map_port_queue_stats_mapping_registers()'
and its sub functions are redundant for majority of drivers.
6) The display of the queue stats and queue stats mapping is mixed
together.
Since xstats is used to obtain queue statistics, we have made the
following simplifications and adjustments:
1) If PMD requires and supports queue stats mapping, configure to driver
in real time by calling ethdev API after executing the command "set
stat_qmap rx/tx ...". If not, the command can not be accepted.
2) Based on the above adjustments, these record structures,
'map_port_queue_stats_mapping_registers()' and its sub functions can
be removed. "tx-queue-stats-mapping" & "rx-queue-stats-mapping"
parameters, and 'parse_queue_stats_mapping_config()' can be removed
too.
3) remove display of queue stats mapping in 'fwd_stats_display()' &
'nic_stats_display()', and obtain queue stats by xstats. Since the
record structures are removed, 'nic_stats_mapping_display()' can be
deleted.
Fixes: 4dccdc789bf4 ("app/testpmd: simplify handling of stats mappings error") Fixes: 013af9b6b64f ("app/testpmd: various updates") Fixes: ed30d9b691b2 ("app/testpmd: add stats per queue") Cc: stable@dpdk.org Signed-off-by: Huisong Li <lihuisong@huawei.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com>
RongQing Li [Wed, 25 Nov 2020 11:01:32 +0000 (19:01 +0800)]
net/af_xdp: optimize Rx mbuf allocation
While receiving packets, the max bunch number of mbufs are allocated
and if hardware does not receive the max bunch number packets, it
will free redundancy mbufs, this is low performance.
So optimize Rx performance, by allocating number of mbuf based on
result of xsk_ring_cons__peek, to avoid to redundancy allocation,
and free mbuf when receive packets.
And Rx cached_cons must be roll backed if fails to allocate mbuf.
Signed-off-by: RongQing Li <lirongqing@baidu.com> Signed-off-by: Dongsheng Rong <rongdongsheng@baidu.com> Acked-by: Ciara Loftus <ciara.loftus@intel.com>
net/bonding: fix port id validity check on parsing
If the port_id is equal to RTE_MAX_ETHPORTS, it should be considered
invalid. Additionally, UNUSED ports are also not valid port ids to be
used afterward.
To simplify following the ethdev API rules, use the exposed function
checking whether a port id is valid.
Fixes: 2efb58cbab6e ("bond: new link bonding library") Cc: stable@dpdk.org Signed-off-by: Gaetan Rivet <grive@u256.net> Acked-by: Min Hu (Connor) <humin29@huawei.com>
Joyce Kong [Mon, 21 Dec 2020 07:38:48 +0000 (15:38 +0800)]
rcu: use EAL memory barrier API
Use rte_atomic_thread_fence wrapper which has been provided for
__atomic_thread_fence builtins to support optimized code for
__ATOMIC_SEQ_CST memory order on x86 platforms.
Signed-off-by: Joyce Kong <joyce.kong@arm.com> Reviewed-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
mbuf: add C++ include guard for dynamic fields header
The header was missing the extern "C" directive which causes name
mangling of functions by C++ compilers, leading to linker errors
complaining of undefined references to these functions.
Fixes: 4958ca3a443a ("mbuf: support dynamic fields and flags") Cc: stable@dpdk.org Signed-off-by: Ashish Sadanandan <ashish.sadanandan@gmail.com> Acked-by: Olivier Matz <olivier.matz@6wind.com>
Ferruh Yigit [Thu, 19 Nov 2020 11:58:57 +0000 (11:58 +0000)]
net/af_xdp: remove useless assignment
Assignment of function parameter 'umem' removed.
Fixes: f0ce7af0e182 ("net/af_xdp: remove resources when port is closed") Cc: stable@dpdk.org Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com> Reviewed-by: David Marchand <david.marchand@redhat.com>
Ferruh Yigit [Thu, 19 Nov 2020 11:58:53 +0000 (11:58 +0000)]
app/procinfo: remove useless memset
The intention with the "sizeof(0)" usage is not clear, but the 'stats'
already 'memset' by 'rte_cryptodev_stats_get()' API, removing 'memset'
in application.
Ferruh Yigit [Thu, 19 Nov 2020 11:58:52 +0000 (11:58 +0000)]
app/procinfo: fix check on xstats-ids
'parse_xstats_ids()' return 'int'. The return value is assigned to
'nb_xstats_ids' unsigned value, later negative check on this variable is
wrong.
Adding interim 'int' variable for negative check.
Fixes: 7ac16a3660c0 ("app/proc-info: support xstats by ID and by name") Cc: stable@dpdk.org Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com> Reviewed-by: David Marchand <david.marchand@redhat.com>
Wisam Jaddo [Thu, 26 Nov 2020 11:15:42 +0000 (13:15 +0200)]
app/flow-perf: change clock measurement
The clock() function is not good practice to use for multiple
cores/threads, since it measures the CPU time used by the process
and not the wall clock time, while when running through multiple
cores/threads simultaneously, we can burn through CPU time much
faster.
As a result this commit will change the way of measurement to use
rd_tsc, and the results will be divided by the processor frequency.
Wisam Jaddo [Thu, 26 Nov 2020 11:15:41 +0000 (13:15 +0200)]
app/flow-perf: add multi-core rule insertion and deletion
One of the ways to increase the insertion/deletion rate is to use
multi-threaded insertion/deletion. Thus it's needed to have support
for testing and measure those rates using flow-perf application.
Now we generate cores and distribute all flows to those cores,
and start inserting/deleting in parallel.
The app now receive the cores count to use from command line option,
then it distribute the rte_flow rules evenly between the cores, and
start inserting/deleting. Each worker will report it's own results,
and in the end the MAIN worker will report the total results for all
cores.
The total results are calculated using RULES_COUNT divided over
max time used between all cores.
Also this touches the memory area, since inserting using multiple cores
in same time the pre solution for memory is not valid, thus now we save
memory before and after each allocation for all cores. In the end we
pick the min pre memory and the max post memory from all cores.
The difference between those values represent the total memory consumed
by the total rte_flow rules from all cores, and then report the total
size of single rte_flow in byte for each port.
Wisam Jaddo [Thu, 26 Nov 2020 11:15:40 +0000 (13:15 +0200)]
app/flow-perf: refactor flows handler
Provide the flows_handler() function the ability to control
flow performance processes. It is made possible after the
introduction of the insert_flows() function.
Also provide to the flows_handler() function the ability to print
the DPDK layer memory consumption of rte_flow rule, regardless
if deletion feature is enabled or not, while in previous
solution it was printing all memory changes after flows_handler().
Thus if deletion is there, it will not provide any memory that
represents the rte_flow rule size.
Also current design is easier to read and understand.
This happens because the dpdk symbols are not exported. Add them to the
dynamic symbol table by using '-Wl,--export-dynamic'. This option was
previously present when compiled with Makefiles, it was introduced in
commit f9a08f650211 ("eal: add support for shared object drivers")
Also add it to the pkg-config file.
Fixes: 16ade738fd0d ("app/testpmd: build with meson") Fixes: 89f0711f9ddf ("examples: build some samples with meson") Cc: stable@dpdk.org Signed-off-by: Olivier Matz <olivier.matz@6wind.com> Reviewed-by: Bruce Richardson <bruce.richardson@intel.com>
Tal Shnaiderman [Sun, 13 Dec 2020 14:16:04 +0000 (16:16 +0200)]
bus/pci: ignore missing NUMA node on Windows
On older processors, NUMA isn't bound to PCIe locality.
those cases return ERROR_NOT_FOUND in response to the
SetupDiGetDevicePropertyW call with DEVPKEY_Device_Numa_Node
attribute.
This error fails the probe process for the PCIe device.
this commit will ignore such failure and will set the
numa_node to 0.
Fixes: b762221ac24f ("bus/pci: support Windows with bifurcated drivers") Cc: stable@dpdk.org Reported-by: Odi Assli <odia@nvidia.com> Signed-off-by: Tal Shnaiderman <talshn@nvidia.com> Tested-by: Odi Assli <odia@nvidia.com> Acked-by: Dmitry Kozlyuk <dmitry.kozliuk@gmail.com> Acked-by: Ranjit Menon <ranjit.menon@intel.com>
Dmitry Kozlyuk [Fri, 11 Dec 2020 20:09:30 +0000 (23:09 +0300)]
bus/pci: fix hardware ID limit on Windows
Length of hardware IDs list is limited by REGSTR_VAL_MAX_HCID_LEN [1],
which is currently 1024. With the old limit of 260, obtaining the list
could fail in a rare occasion of a very long result (no examples known).
This also removes a bogus dependency on the maximum path length.
According to GCC documentation for __builtin_clz:
Returns the number of leading 0-bits in x,
starting at the most significant bit position.
If x is 0, the result is undefined.
__builtin_clz will be called with 0 if the existing
prefix address matches the one we want to insert.
Fixes: 5a5793a5ffa2 ("rib: add RIB library") Cc: stable@dpdk.org Reported-by: David Marchand <david.marchand@redhat.com> Signed-off-by: Vladimir Medvedkin <vladimir.medvedkin@intel.com>
David Marchand [Fri, 4 Dec 2020 17:36:22 +0000 (18:36 +0100)]
ci: enable v21 ABI checks
v21 ABI will be maintained until v21.11.
Let's use the latest released libabigail 1.8.
In GitHub Actions, libabigail binaries and the ABI reference are stored
in two shared caches as all branches can use the same.
While at it, we can reproduce changes from the commit 0b8086ce3fe7
("devtools: remove useless files from ABI reference").
This will save some space in the CI caches.
Signed-off-by: David Marchand <david.marchand@redhat.com> Tested-by: Ferruh Yigit <ferruh.yigit@intel.com> Acked-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Aaron Conole <aconole@redhat.com>
David Marchand [Fri, 4 Dec 2020 17:36:21 +0000 (18:36 +0100)]
ci: hook to GitHub Actions
With the recent changes in terms of free access to the Travis CI, let's
offer an alternative with GitHub Actions.
Running jobs on ARM is not supported unless using external runners, so
this commit only adds builds for x86_64 and cross compiling for i386 and
aarch64.
Differences with the Travis CI integration:
- Error logs are not dumped to the console when something goes wrong.
Instead, they are gathered in a "catch-all" step and attached as
artifacts.
- A cache entry is stored once and for all, but if no cache is found you
can inherit from the default branch cache. The cache is 5GB large, for
the whole git repository.
- The maximum retention of logs and artifacts is 3 months.
- /home/runner is world writable, so a workaround has been added for
starting dpdk processes.
- Ilya, working on OVS GHA support, noticed that jobs can run with
processors that don't have the same capabilities. For DPDK, this
impacts the ccache content since everything was built with
-march=native so far, and we will end up with binaries that can't run
in a later build. The problem has not been seen in Travis CI (?) but
it is safer to use a fixed "-Dmachine=default" in any case.
- Scheduling jobs is part of the configuration and takes the form of a
crontab. A build is scheduled every Monday at 0:00 (UTC) to provide a
default ccache for the week (useful for the ovsrobot).
Signed-off-by: David Marchand <david.marchand@redhat.com> Tested-by: Ferruh Yigit <ferruh.yigit@intel.com> Acked-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Aaron Conole <aconole@redhat.com>
John McNamara [Wed, 2 Dec 2020 19:01:40 +0000 (19:01 +0000)]
license: add licenses for exception cases
The license/exceptions.txt file lists a small number of files
that have licenses that are exceptions to the three main
licenses defined in the Intellectual Property Policy of the
DPDK Charter.
The three exception licenses are MIT, ISC and BSD-2-Clause.
Each of these licenses states that the content of the license
message should be included in the code distribution. This
change adds the text of the MIT, ISC and BSD-2-Clause licenses
as defined on the SPDX website.
Cc: stable@dpdk.org Signed-off-by: John McNamara <john.mcnamara@intel.com> Acked-by: Bruce Richardson <bruce.richardson@intel.com>
Nick Connolly [Sun, 29 Nov 2020 16:00:24 +0000 (16:00 +0000)]
build: disable Windows warnings for insecure funtions
Microsoft CRT defines Windows-specific secure alternatives to
standard library functions and triggers warnings when "insecure"
functions are used [1]. However, calling code already has all
necessary checks around those functions, so these warnings are not
useful for DPDK. MinGW provides its own CRT without this issue.
Nick Connolly [Fri, 27 Nov 2020 11:07:26 +0000 (11:07 +0000)]
eal/windows: fix debug build with MinGW
Compiling with MinGW in --buildtype=debug produces a redefinition
error for strncasecmp.
The root cause is that rte_os.h shouldn't be injecting POSIX definitions
into the environment. It is the applications responsibility to decide
how to handle missing functionality.
Resolving this properly will require further work, but in the meantime
wrap all such definitions with #ifndef/#endif. This resolves the specific
issue with strncasecmp and handles similar issues that applications may
encounter.
Fixes: e8428a9d89f1 ("eal/windows: add some basic functions and macros") Cc: stable@dpdk.org Reported-by: David Marchand <david.marchand@redhat.com> Signed-off-by: Nick Connolly <nick.connolly@mayadata.io>
Dmitry Kozlyuk [Tue, 1 Dec 2020 16:39:59 +0000 (19:39 +0300)]
bus/pci: fix build with MinGW-w64 8
Fix redefinition of GUID, missing from previous versions.
Fixes: b762221ac24f ("bus/pci: support Windows with bifurcated drivers") Cc: stable@dpdk.org Signed-off-by: Dmitry Kozlyuk <dmitry.kozliuk@gmail.com> Tested-by: Thomas Monjalon <thomas@monjalon.net>
Dmitry Kozlyuk [Tue, 1 Dec 2020 16:39:58 +0000 (19:39 +0300)]
eal/windows: fix build with MinGW-w64 8
MinGW-w64 above 8.0.0 exposes VirtualAlloc2() API in headers, but lacks
it in import libraries. Hence, availability of this API at compile-time
can't be used to choose between locating VirtualAlloc2() manually or
relying on the dynamic linker.
Fix redefinition compile-time errors.
Always link VirtualAlloc2() when using GCC.
ethdev: deprecate shared counters using action attribute
A new generic shared actions API may be used to create shared
counter. There is no point to keep duplicate COUNT action specific
capability to create shared counters.
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com> Acked-by: Ori Kam <orika@nvidia.com> Acked-by: Bruce Richardson <bruce.richardson@intel.com>
Anatoly Burakov [Thu, 19 Nov 2020 11:32:32 +0000 (11:32 +0000)]
doc: add VFIO troubleshooting in Linux guide
There are common problems with VFIO that get asked over and over on the
mailing list. Document common problems with VFIO and how to fix them or
at least figure out what went wrong.
Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com> Acked-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: Kevin Traynor <ktraynor@redhat.com>
Anatoly Burakov [Thu, 19 Nov 2020 11:32:29 +0000 (11:32 +0000)]
doc: move VFIO driver to be first in Linux guide
Currently, the Linux GSG mentions UIO drivers first. This is not ideal
as for the longest time, the recommended way to use DPDK with hardware
devices has been to use VFIO driver.
This commit simply moves UIO section after VFIO, with minor edits.
David Marchand [Fri, 27 Nov 2020 15:07:10 +0000 (16:07 +0100)]
usertools: remove dpdk-setup.sh
This old script relied on deprecated stuff, and especially make.
It also applied some scary 666 permissions on files under /dev/vfio.
Its deprecation had been notified in a previous release, remove it.
Signed-off-by: David Marchand <david.marchand@redhat.com> Acked-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Ferruh Yigit <ferruh.yigit@intel.com> Acked-by: Maxime Coquelin <maxime.coquelin@redhat.com>
Ruifeng Wang [Fri, 27 Nov 2020 08:09:03 +0000 (16:09 +0800)]
net/igc: fix build with gcc optimization level 0
GCC build with '-O0' failed for:
../drivers/net/igc/base/igc_api.c
Assembler messages:
29: Error: selected processor does not support `casp x0,x1,x2,x3,[x4]'
82: Error: selected processor does not support `caspa x0,x1,x2,x3,[x4]'
135: Error: selected processor does not support `caspl x0,x1,x2,x3,[x4]'
188: Error: selected processor does not support `caspal x0,x1,x2,x3,[x4]'
This is due to c_args not been passed to meson.
Fixes: 8cb7c57d9b3c ("net/igc: support device initialization") Cc: stable@dpdk.org Reported-by: Feifei Wang <feifei.wang2@arm.com> Signed-off-by: Ruifeng Wang <ruifeng.wang@arm.com> Tested-by: Jerin Jacob <jerinj@marvell.com> Acked-by: Haiyue Wang <haiyue.wang@intel.com>
Ruifeng Wang [Fri, 27 Nov 2020 08:09:02 +0000 (16:09 +0800)]
eal/arm: fix build with gcc optimization level 0
GCC build with '-O0' on platforms with RTE_ARM_FEATURE_ATOMICS set
failed for:
../lib/librte_efd/rte_efd.c
Assembler messages:
3866: Error: selected processor does not support `crc32cb w0,w0,w1'
3890: Error: selected processor does not support `crc32ch w0,w0,w1'
3914: Error: selected processor does not support `crc32cw w0,w0,w1'
3938: Error: selected processor does not support `crc32cx w0,w0,x1'
This was caused by an architecture specifier added for Clang.
Unlike Clang, GCC considers each inline assembly block to be dependent
and therefore, the architecture specifier impacts assemble of some
blocks require certain extension support.
Removed the architecture for GCC to fix the issue.
Fixes: 8fce34cd0a6a ("eal/arm: fix clang build of native target") Cc: stable@dpdk.org Reported-by: Feifei Wang <feifei.wang2@arm.com> Signed-off-by: Ruifeng Wang <ruifeng.wang@arm.com> Acked-by: Jerin Jacob <jerinj@marvell.com>
Anatoly Burakov [Thu, 19 Nov 2020 10:52:45 +0000 (10:52 +0000)]
doc: update information on using hugepages
Current information regarding hugepage usage is a little out of date.
Update it to include information on in-memory mode, as well as on
default mountpoints provided by systemd.
Cc: stable@dpdk.org Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com> Acked-by: Bruce Richardson <bruce.richardson@intel.com>
Anatoly Burakov [Thu, 19 Nov 2020 10:52:44 +0000 (10:52 +0000)]
doc: clarify instructions on running as non-root
The current instructions are slightly out of date when it comes to
providing information about setting up the system for using DPDK as
non-root, so update them.
Gregory Etelson [Thu, 26 Nov 2020 16:43:28 +0000 (17:43 +0100)]
doc: add pkg-config requirement for applications
DPDK relies on pkg-config(1) to provide correct parameters for
compiler and linker used in application build. Inaccurate build
parameters, produced by pkg-config from DPDK .pc files could fail
application build or cause unpredicted results during application
runtime.
Update system requirements doc about a bug in pkg-config v0.27
used in RHEL-7.
Signed-off-by: Gregory Etelson <getelson@nvidia.com> Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Bruce Richardson <bruce.richardson@intel.com>
Thomas Monjalon [Thu, 26 Nov 2020 16:08:37 +0000 (17:08 +0100)]
doc: remove Linux headers from requirements
The compilation of the kernel module KNI is optional.
The kernel headers should not be required for DPDK compilation.
Fixes: 91a861e54164 ("config: disable Linux kernel modules by default") Cc: stable@dpdk.org Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Bruce Richardson <bruce.richardson@intel.com>
Olivier Matz [Thu, 26 Nov 2020 14:23:45 +0000 (15:23 +0100)]
kni: fix build on RHEL 8.3
Like what was done for mainline kernel in commit 38ad54f3bc76 ("kni: fix
build with Linux 5.6"), a new parameter 'txqueue' has to be added to
'ndo_tx_timeout' ndo on RHEL 8.3 kernel.
Asaf Penso [Sun, 22 Nov 2020 12:05:55 +0000 (12:05 +0000)]
doc: update hardware offloads support in mlx5 guide
In DPDK 20.11 the following offload features are added:
* Buffer Split
* Sampling
* Tunnel offload
* 2-port hairpin
* RSS shared action
* Age shared action
Update the relevant tables with OFED/rdma-core/NIC versions.
Ajit Khaparde [Tue, 24 Nov 2020 14:25:03 +0000 (06:25 -0800)]
doc: remove list of supported OS from bnxt guide
Remove list of supported OS in PMD specific doc.
Documenting an unsupported version of OS makes more sense in
PMD specific docs.
Platforms tested with this device is documented in release notes anyway.
doc: add SPDX license tag header to Intel performance guide
This document never had any license or copyright on this file, add one.
Fixes: b932ebcb2664 ("doc: add NIC performance guide on Linux IA") Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: Hemant Agrawal <hemant.agrawal@nxp.com> Acked-by: John McNamara <john.mcnamara@intel.com>
Gregory Etelson [Wed, 18 Nov 2020 16:15:20 +0000 (18:15 +0200)]
doc: update flow API guide for rule removal on stop
There is a discrepancy between ethdev API and flow rules guide
regarding flow rules maintenance after port stop.
librte_ethdev.h declares that flow rules will not be stored in PMD
after port stop:
>>>>> Quote start
Please note that some configuration is not stored between calls to
rte_eth_dev_stop()/rte_eth_dev_start(). The following configuration
will be retained:
- MTU
- flow control settings
- receive mode configuration (promiscuous mode, all-multicast mode,
hardware checksum mode, RSS/VMDQ settings etc.)
- VLAN filtering configuration
- default MAC address
- MAC addresses supplied to MAC address array
- flow director filtering mode (but not filtering rules)
- NIC queue statistics mappings
<<<< Quote end
PMD cannot always correctly restore flow rules after port stop / port
start because application may alter port configuration after port stop
without PMD knowledge about undergoing changes. Consider the
following scenario:
application configures 2 queues 0 and 1 and creates a flow rule with
'queue index 1' action. After that application stops the port and
removes queue 1.
Although PMD can implement flow rule shadow copy to be used for
restore after port start, attempt to restore flow rule from shadow
will fail in example above and PMD could not notify application about
that failure. As the result, flow rules map in HW will differ from
what application expects. In addition, flow rules shadow copy used
for port start restore consumes considerable amount of system memory,
especially in systems with millions of flow rules.
Signed-off-by: Gregory Etelson <getelson@nvidia.com> Acked-by: Ori Kam <orika@nvidia.com> Acked-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Timothy Redaelli [Tue, 24 Nov 2020 15:14:15 +0000 (16:14 +0100)]
eal: fix plugin loading
Commit 49b536fc3060 ("eal: load only shared libs from driver plugin directories")
introduced a check that any shared library must ends with .so, but it can't
work, at least, on Fedora/CentOS/RHEL since .so symlinks are not installed
when you install dpdk package, but only when you install dpdk-devel package.
This commit adds also a check for .so.ABI_VERSION to check for shared lib.
See Fedora Packaging Guidelines for more information:
https://docs.fedoraproject.org/en-US/packaging-guidelines/#_devel_packages
Fixes: 49b536fc3060 ("eal: load only shared libs from driver plugin directories") Cc: stable@dpdk.org Signed-off-by: Timothy Redaelli <tredaelli@redhat.com> Acked-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: David Marchand <david.marchand@redhat.com> Acked-by: Maxime Coquelin <maxime.coquelin@redhat.com> Tested-by: Ferruh Yigit <ferruh.yigit@intel.com> Tested-by: Ali Alnubani <alialnu@nvidia.com>
Timothy Redaelli [Tue, 24 Nov 2020 15:14:14 +0000 (16:14 +0100)]
eal: fix shared lib mode detection
Commit 06c7871dde01 ("eal: restrict default plugin path to shared lib mode")
introduced a check that enabled shared lib mode when librte_eal.so can
be loaded, but it can't work, at least, on Fedora/CentOS/RHEL since .so
symlinks are not installed when you install dpdk package, but only when
you install dpdk-devel package.
This commit uses librte_eal.so.ABI_VERSION to check for shared lib,
since it exists on any linux distributions.
See Fedora Packaging Guidelines for more information:
https://docs.fedoraproject.org/en-US/packaging-guidelines/#_devel_packages
Fixes: 06c7871dde01 ("eal: restrict default plugin path to shared lib mode") Cc: stable@dpdk.org Signed-off-by: Timothy Redaelli <tredaelli@redhat.com> Acked-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: David Marchand <david.marchand@redhat.com> Acked-by: Maxime Coquelin <maxime.coquelin@redhat.com> Tested-by: Ali Alnubani <alialnu@nvidia.com>
git.droids-corp.org / dpdk.git / log