Olivier Matz [Wed, 8 Jan 2020 15:19:15 +0000 (16:19 +0100)]
mempool: fix slow allocation of large mempools
When allocating a mempool which is larger than the largest
available area, it can take a lot of time:
a- the mempool calculate the required memory size, and tries
to allocate it, it fails
b- then it tries to allocate the largest available area (this
does not request new huge pages)
c- add this zone to the mempool, this triggers the allocation
of a mem hdr, which request a new huge page
d- back to a- until mempool is populated or until there is no
more memory
This can take a lot of time to finally fail (several minutes): in step
a- it takes all available hugepages on the system, then release them
after it fails.
The problem appeared with commit eba11e364614 ("mempool: reduce wasted
space on populate"), because smaller chunks are now allowed. Previously,
it had to be at least one page size, which is not the case in step b-.
To fix this, implement our own way to allocate the largest available
area instead of using the feature from memzone: if an allocation fails,
try to divide the size by 2 and retry. When the requested size falls
below min_chunk_size, stop and return an error.
Fixes: eba11e364614 ("mempool: reduce wasted space on populate") Cc: stable@dpdk.org Signed-off-by: Olivier Matz <olivier.matz@6wind.com>
Olivier Matz [Wed, 8 Jan 2020 15:05:41 +0000 (16:05 +0100)]
mempool: fix mempool virt populate with small chunks
To populate a mempool with a virtual area, the mempool code calls
rte_mempool_populate_iova() for each iova-contiguous area. It happens
(rarely) that this area is too small to store one object. In this case,
rte_mempool_populate_iova() returns an error, which is forwarded by
rte_mempool_populate_virt().
This case should not throw an error in
rte_mempool_populate_virt(). Instead, the area that is too small should
just be ignored.
To fix this issue, change the return value of
rte_mempool_populate_iova() to -ENOBUFS when no object can be populated,
so it can be ignored by the caller. As this would be an API change, add
a compat wrapper to keep the current API unchanged. The wrapper will be
removed for 20.11.
David Marchand [Thu, 19 Dec 2019 21:45:03 +0000 (22:45 +0100)]
add ABI checks
Starting from Kevin and Bruce idea of using libabigail, here is an
alternate approach to implement ABI checks.
By default, those checks are disabled and enabling them requires a
manual step that generates the ABI dumps on a reference version for a
set of configurations.
Those checks are enabled in the CI by default for the default meson
options on x86 and aarch64 so that proposed patches are validated.
A cache of the ABI is stored in travis jobs.
Checks can be only informational by setting ABI_CHECKS_WARN_ONLY when
breaking the ABI in a future release.
For advanced developers and maintainers, the contributing guide details
the higher level scripts that are quite close to the existing devtools
scripts.
Signed-off-by: David Marchand <david.marchand@redhat.com>
Bruce Richardson [Wed, 23 Oct 2019 01:07:49 +0000 (01:07 +0000)]
build: enable debug info by default in meson builds
We can turn on debug info by default in meson builds, since it has no
performance penalty. This is done by changing the default build type from
"release" to "debugoptimized". Since the latter using O2, we can using
extra cflags to override that back to O3, which will make little real
difference for actual debugging.
For real debug builds, the user can still do "meson --buildtype=debug ..."
and to remove the debug info "meson --buildtype=release ..." can be used.
These are all standard meson options.
The advantage of having debug builds by default using meson settings is
that we can then add checks for ABI compatibility into each build, and
disable them if we detect that the user has turned off the debug info.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Bruce Richardson [Fri, 10 Jan 2020 21:52:04 +0000 (21:52 +0000)]
doc: rebuild with meson whenever a file changes
Add proper support for calling sphinx whenever a file in the doc
directory changes. This is accomplished by using a wrapper script
for sphinx, which runs sphinx but also emits a gcc-format dependency
file listing all the doc files. This is used by ninja so that any
change to the doc files triggers a rebuild of the docs.
Bruce Richardson [Fri, 10 Jan 2020 21:52:02 +0000 (21:52 +0000)]
doc: fix warning with meson
The install parameter to configure_file is new in 0.50 and generates a
warning since it is newer than our minimum version of 0.47.1. The
parameter, however, is unneeded as the documentation states:
"When omitted it defaults to true when install_dir is set and not empty,
false otherwise."
Given that install_dir is not set for this file, install defaults to false
so no need to explicitly specify it.
Fixes: 720b14db3ae2 ("build: generate API documentation with meson") Cc: stable@dpdk.org Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: Aaron Conole <aconole@redhat.com> Acked-by: Luca Boccassi <bluca@debian.org>
Bruce Richardson [Fri, 10 Jan 2020 21:52:00 +0000 (21:52 +0000)]
kni: fix meson warning about console keyword
Since kni no longer includes the ethtool code and so is faster to build, we
no longer need the console parameter to have incremental screen updates as
it builds. Therefore, we drop the keyword which removes the warning.
Fixes: b78f32cff94d ("kni: support meson build") Cc: stable@dpdk.org Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: Aaron Conole <aconole@redhat.com> Acked-by: Luca Boccassi <bluca@debian.org>
Ruifeng Wang [Mon, 13 Jan 2020 06:26:06 +0000 (14:26 +0800)]
ci: add travis ci support for native aarch64
Add Travis compilation jobs for native aarch64. gcc/clang compilations
for static/shared libraries are added.
Some limitations for current aarch64 Travis support:
1. Container is used. Huge page is not available due to security reason.
2. Missing kernel header package in Xenial distribution.
Solutions to address the limitations:
1. Not to add unit test for now. And run tests with no-huge in future.
2. Use Bionic distribution for all aarch64 jobs.
Signed-off-by: Ruifeng Wang <ruifeng.wang@arm.com> Reviewed-by: Gavin Hu <gavin.hu@arm.com> Acked-by: Aaron Conole <aconole@redhat.com>
Ruifeng Wang [Mon, 13 Jan 2020 06:26:05 +0000 (14:26 +0800)]
devtools: add path to additional shared object files
Drivers librte_mempool_ring.so and librte_pmd_null.so are loaded by
librte_eal.so when running testpmd.
In Ubuntu Xenial, driver path is installed to RPATH on testpmd. This
allows librte_eal.so to find drivers by using the RPATH.
However, in Ubuntu Bionic, driver path is installed to RUNPATH instead.
The RUNPATH on testpmd is not available by librte_eal.so and therefore
lead to driver load failure:
EAL: Detected 32 lcore(s)
EAL: Detected 1 NUMA nodes
EAL: librte_mempool_ring.so: cannot open shared object file:
No such file or directory
EAL: FATAL: Cannot init plugins
EAL: Cannot init plugins
Add 'drivers' into LD_LIBRARY_PATH so that testpmd can find and make
use of these shared libraries.
Signed-off-by: Ruifeng Wang <ruifeng.wang@arm.com> Reviewed-by: Gavin Hu <gavin.hu@arm.com> Acked-by: Aaron Conole <aconole@redhat.com>
David Marchand [Thu, 9 Jan 2020 12:59:16 +0000 (13:59 +0100)]
ci: use meson 0.47.1
meson 0.53.0 has a compatibility issue [1] with the python 3.5.2 that comes
in Ubuntu 16.04.
On the other hand, the minimal version supported in dpdk is 0.47.1.
Stick to this version to avoid getting hit by regressions in meson latest
shiny release.
David Marchand [Thu, 9 Jan 2020 12:59:15 +0000 (13:59 +0100)]
build: fix libm detection in meson
Using version 0.47.1, meson is unable to find the math library in Travis
for the 32bits job.
Quite surprisingly, this problem is not seen with the 64bits jobs.
Switching to 0.48.0, the problem disappears.
But we should pass 'm' to find_library instead of 'libm' anyway.
Fixes: 98edcbb5ab2f ("eal/windows: introduce Windows support") Cc: stable@dpdk.org Signed-off-by: David Marchand <david.marchand@redhat.com> Acked-by: Bruce Richardson <bruce.richardson@intel.com>
Bruce Richardson [Mon, 23 Dec 2019 10:38:40 +0000 (10:38 +0000)]
build: explicitly enable sse4 for meson
If the compiler does not recognise the specific CPU when building with the
default "native" machine type, sse4.2 instructions can be missing, causing
a build error. Rather than advising the user to change the machine type,
we can just turn on SSE4.2 directly. This can prevent issues with running
automated tests with older compilers/distros on newer hardware.
Cc: stable@dpdk.org Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: David Marchand <david.marchand@redhat.com>
Aaron Conole [Tue, 3 Dec 2019 21:15:44 +0000 (16:15 -0500)]
service: don't walk out of bounds when checking services
The service_valid call is used without properly bounds checking the
input parameter. Almost all instances of the service_valid call are
inside a for() loop that prevents excessive walks, but some of the
public APIs don't bounds check and will pass invalid arguments.
Prevent this by using SERVICE_GET_OR_ERR_RET where it makes sense,
and adding a bounds check to one service_valid() use.
Fixes: 8d39d3e237c2 ("service: fix race in service on app lcore function") Fixes: e9139a32f6e8 ("service: add function to run on app lcore") Fixes: e30dd31847d2 ("service: add mechanism for quiescing") Cc: stable@dpdk.org Signed-off-by: Aaron Conole <aconole@redhat.com> Reviewed-by: David Marchand <david.marchand@redhat.com>
Wrong rte_log2_u64(0) val 0, expected ffffffff
Test Failed
The ut passes 0 to log2() to get an expected value.
Quoting log2 / log(3) manual:
If x is zero, then a pole error occurs, and the functions return
-HUGE_VAL, -HUGE_VALF, or -HUGE_VALL, respectively.
rte_log2_uXX helpers handle 0 as a special value and return 0.
Let's have dedicated tests for this case.
Fixes: 05c4345ef5c2 ("test: add unit test for integer log2 function") Cc: stable@dpdk.org Signed-off-by: David Marchand <david.marchand@redhat.com> Acked-by: Aaron Conole <aconole@redhat.com>
Bruce Richardson [Thu, 12 Dec 2019 11:58:26 +0000 (11:58 +0000)]
build: fix soname info for 19.11 compatibility
The soname for each stable ABI version should be just the ABI version major
number without the minor number. Unfortunately both major and minor were
used causing version 20.1 to be incompatible with 20.0.
This patch fixes the issue by switching from 2-part to 3-part ABI version
numbers so that we can keep 20.0 as soname and using the final digits to
identify the 20.x releases which are ABI compatible. This requires changes
to both make and meson builds to handle the three-digit version and shrink
it to 2-digit for soname.
The final fix needed in this patch is to adjust the library version number
for the ethtool example library, which needs to be upped to 2-digits, as
external libraries using the DPDK build system also use the logic in this
file.
Fixes: cba806e07d6f ("build: change ABI versioning to global") Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: Neil Horman <nhorman@tuxdriver.com> Tested-by: Ray Kinsella <mdr@ashroe.eu> Tested-by: Ferruh Yigit <ferruh.yigit@intel.com> Tested-by: Kevin Laatz <kevin.laatz@intel.com> Tested-by: David Marchand <david.marchand@redhat.com>
Thomas Monjalon [Mon, 16 Dec 2019 14:45:32 +0000 (15:45 +0100)]
devtools: fix debug build test
When testing build with +debug options, the statistics are enabled.
It was wrongly matching CONFIG_RTE_IBVERBS_LINK_STATIC.
The pattern is fixed to match only statistics config options.
Fixes: 2c0dd7b69fb0 ("config: add static linkage of mlx dependency") Cc: stable@dpdk.org Reported-by: Ali Alnubani <alialnu@mellanox.com> Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
Ali Alnubani [Wed, 11 Dec 2019 06:49:25 +0000 (06:49 +0000)]
net/mlx: fix build with clang 9
This rewrites the MKSTR macro appending an empty string to its arguments
to resolve build failures similar to:
drivers/net/mlx4/mlx4.c:461:14: fatal error: format string is not a
string literal [-Wformat-nonliteral]
MKSTR(path, "%s/device/uevent", device->ibdev_path);
drivers/net/mlx5/mlx5_stats.c:144:15: fatal error: format string is not a
string literal [-Wformat-nonliteral]
MKSTR(path, "%s/ports/%d/hw_counters/%s",
Thomas Monjalon [Mon, 9 Dec 2019 20:43:12 +0000 (21:43 +0100)]
doc: fix build with python 3.8
After upgrading to python-3.8.0, a syntax mismatch is revealed:
doc/guides/conf.py:240: SyntaxWarning: "is not" with a literal.
Did you mean "!="?
if value is not '':
Removing "is not ''" seems the right thing to do.
A patch may also be needed in the RTD theme package:
https://github.com/readthedocs/sphinx_rtd_theme/commit/a49a812c.diff
(not included in release 0.4.3)
Fixes: 9db3f52126fb ("doc: generate NIC overview table from ini files") Cc: stable@dpdk.org Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: Bruce Richardson <bruce.richardson@intel.com>
David Marchand [Wed, 4 Dec 2019 13:16:22 +0000 (14:16 +0100)]
eal/linux: fix build error on RHEL 7.6
Previous fix gives hiccups to gcc on RHEL 7.6:
== Build lib/librte_eal/linux/eal
CC eal_interrupts.o
...lib/librte_eal/linux/eal/eal_interrupts.c: In function
‘eal_intr_thread_main’:
...lib/librte_eal/linux/eal/eal_interrupts.c:1048:9: error: missing
initializer for field ‘events’ of ‘struct epoll_event’
[-Werror=missing-field-initializers]
struct epoll_event ev = { };
^
In file included from ...lib/librte_eal/linux/eal/eal_interrupts.c:15:0:
/usr/include/sys/epoll.h:89:12: note: ‘events’ declared here
uint32_t events; /* Epoll events */
^
...lib/librte_eal/linux/eal/eal_interrupts.c: At top level:
cc1: error: unrecognized command line option
"-Wno-address-of-packed-member" [-Werror]
cc1: all warnings being treated as errors
Fixes: e0ab8020ac2a ("eal/linux: fix uninitialized data valgrind warning") Cc: stable@dpdk.org Reported-by: Andrew Rybchenko <arybchenko@solarflare.com> Signed-off-by: David Marchand <david.marchand@redhat.com>
eal/linux: fix uninitialized data valgrind warning
Valgrind reports that eal interrupt thread is calling epoll_ctl
with uninitialized data.
This is a false positive, because the kernel is not going to care about
the unused bits in the union but trivial to fix by initializing it.
Fixes: af75078fece3 ("first public release") Cc: stable@dpdk.org Signed-off-by: Stephen Hemminger <stephen@networkplumber.org> Acked-by: David Marchand <david.marchand@redhat.com>
Ali Alnubani [Thu, 28 Nov 2019 07:24:38 +0000 (07:24 +0000)]
config: add pkgconfig for arm64
Meson fails to find a pkg-config executable if pkgconfig
isn't set for aarch64. The environment variable `PKG_CONFIG_PATH`
is useless in this case, and meson fails to locate dependencies
that are built in non-standard paths.
Signed-off-by: Ali Alnubani <alialnu@mellanox.com> Acked-by: Bruce Richardson <bruce.richardson@intel.com>
Bruce Richardson [Thu, 28 Nov 2019 16:33:50 +0000 (16:33 +0000)]
doc: update example output in FreeBSD guide
The output of running the helloworld example on FreeBSD was a little
out-of-date and can be shortened by using the latest version of DPDK.
Update appropriately.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: Anatoly Burakov <anatoly.burakov@intel.com>
Bruce Richardson [Thu, 28 Nov 2019 16:33:49 +0000 (16:33 +0000)]
doc: remove reference to old version of FreeBSD
FreeBSD 10 is now EOL and all testing with DPDK takes place on BSD versions
11 and 12, so we can just remove the note. The BSD ports are supported on
all non-EOL versions of BSD.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: Anatoly Burakov <anatoly.burakov@intel.com>
Bruce Richardson [Thu, 28 Nov 2019 16:33:48 +0000 (16:33 +0000)]
doc: reorder meson and make build instructions for Arm
Since the meson instructions are the simpler of the two sets, and also the
ones most future-proof, put those first in the user documentation with make
instructions following them.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Acked-by: Gavin Hu <gavin.hu@arm.com>
Bruce Richardson [Thu, 28 Nov 2019 16:33:46 +0000 (16:33 +0000)]
doc: update system requirements in Linux guide
Update the system requirements section of the doc to cover builds with
meson and ninja. This involves updating the package dependencies to include
meson, ninja and python 3.5, and also updating the optional dependencies
section to explain that the components are enabled/disabled automatically
by meson.
As part of this update, the relevant sections were simplified to keep the
document shorter. For mandatory requirements, we can refer to the various
distro's development tools package groups rather than requiring gcc, core
tools etc. individually. The optional package list was very incomplete, and
if complete would duplicate information in the individual driver's guides.
Therefore we can simplify it by listing only the library optional
requirements and referring users to the driver docs to find details on
their dependencies.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Bruce Richardson [Wed, 27 Nov 2019 23:00:55 +0000 (00:00 +0100)]
devtools: control location of test builds
By default, both test-build.sh and test-meson-builds.sh scripts create the
builds they generate in the current working directory, leading to a large
number of build directories being present when testing patches. This
patchset modifies both scripts to use a DPDK_BUILD_TEST_DIR environment
variable to control where the build outputs are put.
For example, doing:
export DPDK_BUILD_TEST_DIR=__builds
./devtools/test-meson-builds.sh && ./devtools/test-build.sh \
x86_64-native-linux-clang+shared i686-native-linux-gcc
gives a "__builds" directory with 14 meson and 2 make builds (with the
meson build count depending on compiler availability)
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com> Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: David Marchand <david.marchand@redhat.com>
Test vectors for AES-GCM need to be populated by using aead member
and not cipher_auth. Using incorrect member would result in failures
when AES-GCM with non-zero AAD is tried.
Bugzilla ID: 360 Fixes: 4aaad2995e13 ("examples/fips_validation: support GCM parsing") Signed-off-by: Anoob Joseph <anoobj@marvell.com> Signed-off-by: Sucharitha Sarananaga <ssarananaga@marvell.com>
Ferruh Yigit [Thu, 28 Nov 2019 12:15:26 +0000 (12:15 +0000)]
kni: fix build with Linux 4.9.x
The 'get_user_pages_remote()' API is updated in kernel 4.10.0 [1],
but the check added as > 4.9.0,
this logic is broken for kernels 4.9.x, because they justify
> 4.9.0 check but have the old API.
Fixing the check as >= 4.10.0
[1]
commit 5b56d49fc31d ("mm: add locked parameter to get_user_pages_remote()")
Fixes: d965af9e8ae1 ("kni: increase kernel version requirement for VA") Reported-by: Andrew Rybchenko <arybchenko@solarflare.com> Suggested-by: David Marchand <david.marchand@redhat.com> Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com> Tested-by: Andrew Rybchenko <arybchenko@solarflare.com> Reviewed-by: David Marchand <david.marchand@redhat.com>
David Marchand [Tue, 8 Oct 2019 15:52:29 +0000 (17:52 +0200)]
devtools: check coverity and bugzilla tags
Let's try to check for discrepancies in coverity and bugzilla tags.
The contributing guide specifies that:
- for coverity issues, the tag is 'Coverity issue:'
- for bugzilla issues, the tag is 'Bugzilla ID:'
Signed-off-by: David Marchand <david.marchand@redhat.com> Acked-by: Kevin Traynor <ktraynor@redhat.com>
David Marchand [Wed, 2 Oct 2019 16:55:47 +0000 (18:55 +0200)]
devtools: reset compilation flags for each target
Same idea than overriding PATH and PKG_CONFIG_PATH, it can be quite
useful to override compilation flags like CFLAGS, CPPFLAGS and LDFLAGS
for cross compilation or libraries that won't provide a pkg-config file.
Thomas Monjalon [Sat, 14 Sep 2019 15:05:09 +0000 (17:05 +0200)]
git: ignore more build directories
Build directories commonly have compiler in their names.
In order to filter build directories not starting with "build-"
(common with make), patterns for gcc and clang are added to .gitignore.
Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Acked-by: David Marchand <david.marchand@redhat.com>
Harry van Haaren [Wed, 27 Nov 2019 13:20:27 +0000 (13:20 +0000)]
test/service: fix wait for service core
This commit fixes a sporadic failure of the service_autotest
unit test, as seen in the DPDK CI. The failure occurs as the main test
thread did not wait on the service-thread to return, and allowing it
to read a flag before the service was able to write to it.
The fix changes the wait API call to specific the service-core ID,
and this waits for cores with both ROLE_RTE and ROLE_SERVICE.
The rte_eal_mp_wait_lcore() call does not (and should not) wait
for service cores, so must not be used to wait on service-cores.
Fixes: f038a81e1c56 ("service: add unit tests") Cc: stable@dpdk.org Reported-by: Aaron Conole <aconole@redhat.com> Signed-off-by: Harry van Haaren <harry.van.haaren@intel.com> Acked-by: David Marchand <david.marchand@redhat.com>
Pavan Nikhilesh [Wed, 27 Nov 2019 12:36:47 +0000 (18:06 +0530)]
event/octeontx: fix partial Rx packet handling
When net/octeontx is connected to event/octeontx as an event Rx adapter,
PKI aka 'net/octeontx' can forward packets directly to SSO aka
'event/octeontx'.
When pumping traffic to PKI if flow control is disabled internal FIFOs
might be overrun causing partial l2 packets to be enqueued.
SSO receives <31:0> TAG tag calculated by PKI, in normal cases <31:28>
is always 0 which signifies RTE_EVENT_TYPE_ETHDEV. But in case of
partial received packets PKI sets the <31:0> TAG as 0xFFFFFFFF which
is an invalid event type.
Add a check to see if TAG is 0xFFFFFFFF and free the partial receive
packet.
Fixes: d0d654986018 ("net/octeontx: support event Rx adapter") Cc: stable@dpdk.org Signed-off-by: Pavan Nikhilesh <pbhagavatula@marvell.com> Acked-by: Jerin Jacob <jerinj@marvell.com>
The legacy multi-packet write is the feature allowing to
put multiple packets into one transmitting descriptor,
this feature is supported by only NIC ConnectX-4LX.
The number of packets should be limited to provide
optimal size descriptor and better performance.
The legacy multi-packet write is the feature allowing to
put multiple packets into one transmitting descriptor,
this feature is supported by only NIC ConnectX-4LX.
The number of packets should be limited to provide
optimal size descriptor and better performance.
In extensive metadata mode the MARK copy table is engaged,
if the application creates the flow with zero MARK ID action:
flow create 1 ingress pattern eth / ... / end actions mark id 0 / .. end
And then destroys that, the traffic to the port stops. This happens
due to default flow for the copy table has the zero ID and is removed
with the application rule. The patch extends internal ID variable
to 64 bits and provide the UINT64_MAX ID for the copy table default
rule.
Fixes: dd3c774f6ffb ("net/mlx5: add metadata register copy table") Signed-off-by: Viacheslav Ovsiienko <viacheslavo@mellanox.com> Acked-by: Ori Kam <orika@mellanox.com>
Thomas Monjalon [Wed, 27 Nov 2019 12:22:56 +0000 (13:22 +0100)]
ethdev: limit maximum number of queues
A buffer overflow happens in testpmd with some drivers
since the queue arrays are limited to RTE_MAX_QUEUES_PER_PORT.
The advertised capabilities of mlx4, mlx5 and softnic
for the number of queues were the maximum number: UINT16_MAX.
They must be limited by the configured RTE_MAX_QUEUES_PER_PORT
that applications expect to be respected.
The limitation is applied at ethdev level (function rte_eth_dev_info_get),
in order to force the configured limit for all drivers.
Fixes: 14b53e27b30e ("ethdev: fix crash with multiprocess") Cc: stable@dpdk.org Reported-by: Raslan Darawsheh <rasland@mellanox.com> Signed-off-by: Thomas Monjalon <thomas@monjalon.net> Reviewed-by: Ferruh Yigit <ferruh.yigit@intel.com> Reviewed-by: David Marchand <david.marchand@redhat.com>
Ray Kinsella [Mon, 18 Nov 2019 13:15:28 +0000 (13:15 +0000)]
doc: add ABI policy changes to release notes
Add some pointers to the releases notes on the changes to the abi policy,
the introduction of project-level ABI management and the deprecation of
library-level management.
Signed-off-by: Ray Kinsella <mdr@ashroe.eu> Acked-by: John McNamara <john.mcnamara@intel.com>
Pavan Nikhilesh [Sat, 2 Nov 2019 14:23:15 +0000 (19:53 +0530)]
usertools: fix device binding module detection
Some kernel modules use '-' in their name when registering through
`pci_register_driver` and the same name is populated in
'/sys/bus/pci/drivers/'.
But the kernel always populates modules names replacing '-' with '_'
in '/sys/module/'.
Example:
# ./usertools/dpdk-devbind.py -b octeontx2-nicpf 0002:03:00.0
Error: Driver 'octeontx2-nicpf' is not loaded.
# ls /sys/bus/pci/drivers/octeontx2-nicpf
bind module new_id remove_id uevent unbind
# ls /sys/module/octeontx2_nicpf/
drivers uevent version
The patch addresses it by always replacing '-' with '_' when looking in
'/sys/module/'
Suanming Mou [Tue, 26 Nov 2019 14:08:35 +0000 (16:08 +0200)]
net/mlx5: fix crash on GRE flow rule parsing
When set the GRE item, GRE key should follow after GRE header, or the
header gre_item pointer used by the key will be invalid.
Currently in the mlx5_flow_validate_item_gre_key() function, the header
gre_item pointer is access before checking if the key is after the
header or not. Once the key item is before the header, invalid gre_item
pointer access happens.
Move the gre_item pointer access after the GRE header check to avoid the
crash issue.
Fixes: a7a0365565a4 ("net/mlx5: match GRE key and present bits") Cc: stable@dpdk.org Signed-off-by: Suanming Mou <suanmingm@mellanox.com> Acked-by: Ori Kam <orika@mellanox.com>
Ali Alnubani [Tue, 26 Nov 2019 08:35:24 +0000 (08:35 +0000)]
net/mlx5: fix build with dlopen linkage
The header mlx5_glue.h includes rte_byteorder.h, which isn't
pulled by global_inc. This patch adds the missing include directory
and fixes the build error:
In file included from drivers/net/mlx5/mlx5_glue.c:31:
drivers/net/mlx5/mlx5_glue.h:12:10: fatal error:
rte_byteorder.h: No such file or directory
The build failure can be triggered with:
meson -Dibverbs_link=dlopen build && ninja -C build
Fixes: 400d985eb586 ("net/mlx5: add VLAN push/pop DR commands to glue") Signed-off-by: Ali Alnubani <alialnu@mellanox.com> Acked-by: Thomas Monjalon <thomas@monjalon.net>
ConnectX-4LX supports multiple packets within the single Tx
descriptor. This feature is named as "Legacy Multi-Packet Write"
and imposes a lot of limitations:
- no ACLs, it means no NIC Tx Flows are supported and Tx metadata
become meaningless
- the required minimal inline data must be zero
- no SR-IOV, it means no support in E-Switch configurations,
- no priority and dscp forcing
- no VLAN insertion
- no TSO
- all packets within MPW session must have the same size
This legacy MPW feature is mainly intended for test purposes.
To explicitly engage the feature on ConnectX-4LX the devargs
should be specified:
- txq_mpw_en=1
This feature was dropped in 19.08, this patch reverts it back.
Ferruh Yigit [Tue, 26 Nov 2019 10:50:07 +0000 (10:50 +0000)]
net/kni: fix crash caused by double stop
'close()' calls 'stop()' and 'stop()' cancels pthread without any check.
Calling 'stop()' & 'close()' sequentially tries to cancel pthread twice
which will cause a crash.
Adding a state check in 'stop()' before canceling the pthread to prevent
multiple stop.
Fixes: 696fbc7bb4fc ("net/kni: remove resources when port is closed") Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com> Tested-by: Yinan Wang <yinan.wang@intel.com>
Matan Azrad [Tue, 26 Nov 2019 09:04:10 +0000 (09:04 +0000)]
ethdev: fix item expansion for RSS flow
When the last item in flow pattern includes "next protocol" field which
is relevant for RSS flow expansion, a new item is added to the pattern
according to the "next protocol" field. This field is called missed
field.
The missed field wrongly was not initialized what caused to some of the
flow item fields to contain garbage values.
As a result, the PMDs internal flow engine may crash.
For example, the spec value may include garbage pointer and to cause
crash.
Initialize the missed field with zeroes.
Fixes: fc2dd8dd492f ("ethdev: fix expand RSS flows") Cc: stable@dpdk.org Signed-off-by: Matan Azrad <matan@mellanox.com> Acked-by: Ori Kam <orika@mellanox.com>
Dekel Peled [Tue, 26 Nov 2019 08:34:22 +0000 (10:34 +0200)]
net/mlx5: fix flow engine choice
Commit in fixes line sets the DV (Direct Verbs) flow engine as default.
Newer versions of DV flow engine use the DR (Direct Rules) features.
DR is supported from RDMA Core library version rdma-core-24.0.
This cause failure to start port when using older rdma-core version,
without DR support.
This patch selects DV flow engine if rdma-core version is v24.0 or
higher. Verbs flow engine is selected otherwise.
Fixes: cd4569d2bf3c ("net/mlx5: change default flow engine to DV") Signed-off-by: Dekel Peled <dekelp@mellanox.com> Acked-by: Ori Kam <orika@mellanox.com>
Lunyuan Cui [Wed, 13 Nov 2019 12:46:59 +0000 (12:46 +0000)]
net/e1000: fix link status
The link status got from link status register was not correct,
because register has been reset when ports reset.
After port reset, set the link status down.
Fixes: c431ec66c54c ("net/igb: support setting link up or down") Cc: stable@dpdk.org Signed-off-by: Lunyuan Cui <lunyuanx.cui@intel.com> Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
Lunyuan Cui [Mon, 18 Nov 2019 15:37:44 +0000 (15:37 +0000)]
net/ixgbe: fix link status
The link status for 82599eb got from link status register was not
correct. Check the enable/disable flag of tx laser, set the link
status down if tx laser disabled. Then, we can get correct status.
But after port reset, tx laser register will be reset enable.
Link status will always be up. So set tx laser disable when port resets.
When hw->mac.autotry_restart is true, whether tx laser is disable or
enable, it will be set enable in ixgbe_flap_tx_laser_multispeed_fiber().
hw->mac.autotry_restart can be set true in both port init and port start.
Because we don't need this treatment before port starts, set
hw->mac.autotry_restart false when port init.
Fixes: 0408f47ba4d6 ("net/ixgbe: fix busy polling while fiber link update") Cc: stable@dpdk.org Signed-off-by: Lunyuan Cui <lunyuanx.cui@intel.com> Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
Lunyuan Cui [Wed, 20 Nov 2019 09:22:03 +0000 (09:22 +0000)]
net/e1000: fix link status update
Meaningless to judge the link state according to the memset'ed
link variable, this patch fixes this logical issue.
In addition, this patch changes the variable from link_check to link_up
according to its real meaning.
Fixes: 80ba61115e77 ("net/e1000: use link status helper functions") Cc: stable@dpdk.org Signed-off-by: Lunyuan Cui <lunyuanx.cui@intel.com> Acked-by: Xiaolong Ye <xiaolong.ye@intel.com> Acked-by: Wenzhuo Lu <wenzhuo.lu@intel.com>
Wei Hu (Xavier) [Mon, 25 Nov 2019 09:00:53 +0000 (17:00 +0800)]
net/hns3: fix checking enough Tx BDs
In .tx_pkt_burst ops implementation function of hns3 PMD driver,
there is one check whether there are enough BDs in the TX queue.
If not, driver will stop sending the packets.
Currently in the 'for' process loop, the next_to_use member of
TX queue is not updated in time after processing BDs of one packet,
which results in the invalid action of checking whether there are
enough BDs and failure in sending packets.
This patch fixes it by moving the assignment statment of the
next_to_use member of TX queue to the place after porcessing TX BDs
in the 'for' loop.
Fixes: bba636698316 ("net/hns3: support Rx/Tx and related operations") Cc: stable@dpdk.org Signed-off-by: Hongbo Zheng <zhenghongbo3@huawei.com> Signed-off-by: Huisong Li <lihuisong@huawei.com> Signed-off-by: Wei Hu (Xavier) <xavier.huwei@huawei.com>
Rasesh Mody [Fri, 22 Nov 2019 23:31:18 +0000 (15:31 -0800)]
doc: update qede guide
- Add note for sharing an adapter between DPDK and Linux drivers.
- Update the firmware version in example.
- Add Config note for potential error due to lack of memzone descriptor
count.
Pavan Nikhilesh [Sat, 23 Nov 2019 08:58:31 +0000 (14:28 +0530)]
net/octeontx2: fix packet type parsing disabled by default
Enable packet type parsing by default, only disable when
`rte_eth_dev_set_ptypes()` is called with ptype_mask as 0.
This would enable applications that are dependent on packet type parsing
like l3fwd.
Also this patch preserves configuration set by `rte_eth_dev_set_ptypes`
across multiple calls of `rte_eth_dev_configure()`
Fixes: d2706e15e6fb ("net/octeontx2: support reduced set of packet types") Signed-off-by: Pavan Nikhilesh <pbhagavatula@marvell.com> Acked-by: Jerin Jacob <jerinj@marvell.com>
Suanming Mou [Thu, 21 Nov 2019 12:09:30 +0000 (14:09 +0200)]
net/mlx5: fix chosen L3/L4 layer with tunnel
For tunnel mode, there may be two L3/L4 layer match pattern items, one
for inner and one for outer layers. The L3 TTL and L4 port header
modify actions should handle the outermost layer items.
Currently flow_dv_attr_init() doesn't distinguish between outer and
inner layers, since inner layer comes later than the outer layer, this
may cause inner layer values also set to the flow attributes and may
lead actions to use inner L3/L4 pattern items.
Adding a check in flow_dv_attr_init() to prevent inner layer to set the
flow attribute if the previous L3/L4 outer pattern exist.
Fixes: 4bb14c83df95 ("net/mlx5: support modify header using Direct Verbs") Cc: stable@dpdk.org Signed-off-by: Suanming Mou <suanmingm@mellanox.com> Acked-by: Matan Azrad <matan@mellanox.com>
git.droids-corp.org / dpdk.git / log