Age | Commit message (Collapse) | Author |
|
|
|
Explicitly disable the netmap pktio type when attempting to test the
socket based types. Without this on systems with netmap support enabled
(the modules loaded) the socket interfaces may not be tested.
Signed-off-by: Stuart Haslam <stuart.haslam@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
Removed odp_sync_stores() implementation and replaced usage with
odp_mb_full(), to minimize functional changes. Added minimal
validation test function for the new memory barriers.
Signed-off-by: Petri Savolainen <petri.savolainen@nokia.com>
Reviewed-by: Bill Fischofer <bill.fischofer@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
Added new memory barriers. These follow C11 release /
acquire specification and replaces odp_sync_stores().
Used GCC __atomic_thread_fence to implement all three
barriers.
Signed-off-by: Petri Savolainen <petri.savolainen@nokia.com>
Reviewed-by: Bill Fischofer <bill.fischofer@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
The linux-generic implementation of odp_time_t makes use of POSIX
APIs that are sensitive to the _POSIX_C_SOURCE level. Use an indirection
mechanism so that these dependencies do not "bleed through" the ODP API.
This means that ODP applications can be independent of _POSIX_C_SOURCE
level.
Signed-off-by: Bill Fischofer <bill.fischofer@linaro.org>
Reviewed-by: Ivan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
Conflicts:
platform/linux-generic/odp_time.c
|
|
Added validation tests for ODP std C library API.
Signed-off-by: Petri Savolainen <petri.savolainen@nokia.com>
Reviewed-and-Tested-by: Bill Fischofer <bill.fischofer@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
Some C library calls are often used in data plane code. This
API enables possibility to HW optimized implementation of those.
Added first memcpy and memset.
Signed-off-by: Petri Savolainen <petri.savolainen@nokia.com>
Reviewed-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
Conflicts:
platform/linux-generic/Makefile.am
|
|
Signed-off-by: Peng <hepeng@ict.ac.cn>
Reviewed-by: Petri Savolainen <petri.savolainen@nokia.com>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
Conflicts:
platform/linux-generic/Makefile.am
test/validation/hash/hash.c
test/validation/hash/hash.h
|
|
Changes drop policy enum name from odp_drop_e to odp_cls_drop_t
Signed-off-by: Balasubramanian Manoharan <bala.manoharan@linaro.org>
Reviewed-by: Petri Savolainen <petri.savolainen@nokia.com>
Reviewed-and-tested-by: Bill Fischofer <bill.fischofer@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
create api
Implements odp_cls_cos_create() and odp_cls_cos_param_init() functions and
replace odp_cos_create() function with odp_cls_cos_create() in validation
and examples.
Class of service create function now takes pool, queue, drop policy and
name as input parameters.
Adds class of service parameter structure odp_cls_cos_param_t and
initialization function odp_cls_cos_param_init()
Signed-off-by: Balasubramanian Manoharan <bala.manoharan@linaro.org>
Reviewed-by: Petri Savolainen <petri.savolainen@nokia.com>
Reviewed-and-tested-by: Bill Fischofer <bill.fischofer@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
This patch adds global time API and implements it in linux-generic.
Signed-off-by: Ivan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Reviewed-by: Petri Savolainen <petri.savolainen@nokia.com>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
This patch adds odp_time_local_res() and odp_time_wait_until(),
odp_time_wait_ns() APIs.
Signed-off-by: Ivan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Reviewed-by: Petri Savolainen <petri.savolainen@nokia.com>
Tested-by: Mike Holmes <mike.holmes@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
Conflicts:
platform/linux-generic/odp_time.c
|
|
The local time API supposes the time source is wall time.
So correct linux-generic implementation.
Reviewed-by: Petri Savolainen <petri.savolainen@nokia.com>
Signed-off-by: Ivan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
Conflicts:
platform/linux-generic/include/odp_internal.h
|
|
The linux-generic time API implementation shouldn't depend on cpu
cycle API wich is not stable enough to measure time period due to
dynamic frequency scaling.
Reviewed-by: Petri Savolainen <petri.savolainen@nokia.com>
Signed-off-by: Ivan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
Conflicts:
platform/linux-generic/Makefile.am
platform/linux-generic/odp_time.c
|
|
In the schedule_loop the wait time is passed in units of scheduler
time. At this moment there is no difference between wait time and
odp_time, but in case if odp_time is smth different from uint64_t,
it cannot be directly passed as uint64_t. So better to pass scheduler
time as ns and convert them before entering a loop.
Reviewed-by: Petri Savolainen <petri.savolainen@nokia.com>
Signed-off-by: Ivan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
Fixes a crash caused by pool not assigned to CoS.
Signed-off-by: Balasubramanian Manoharan <bala.manoharan@linaro.org>
Reviewed-and-Tested-by: Maxim Uvarov <maxim.uvarov@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
Fixes netmap "cast discards '__attribute__((const))'" build
error. Define pkt_addr as const as classifier operations
should be read-only to packet data.
Signed-off-by: Matias Elo <matias.elo@nokia.com>
Reviewed-by: Balasubramanian Manoharan <bala.manoharan@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
Fixes arm build failure due to wrong typecast of 'const' variable
Signed-off-by: Balasubramanian Manoharan <bala.manoharan@linaro.org>
Reviewed-by: Maxim Uvarov <maxim.uvarov@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
Adds support for configuring packet pool to a class-of-service.
linux-generic packet parser is enhanced to parse a packet directly from
a memory location rather than from odp_packet_t.
packet receive code is modified to run packet classifier directly from
the stream so that the packet can be allocated from the pool specified
by the CoS.
Signed-off-by: Balasubramanian Manoharan <bala.manoharan@linaro.org>
Reviewed-by: Petri Savolainen <petri.savolainen@nokia.com>
Reviewed-and-tested-by: Bill Fischofer <bill.fischofer@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
Replaced internal define with ODP_THREAD_COUNT_MAX. This
fixes also the current make install issue.
Signed-off-by: Petri Savolainen <petri.savolainen@nokia.com>
Reviewed-by: Bill Fischofer <bill.fischofer@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
Removed module synchronizer from doxygen documentation and
introduced new modules for locks, atomics and barriers. Removed
unnecessary group tagging from internal headers, which are not
visible to doxygen anyway.
Signed-off-by: Petri Savolainen <petri.savolainen@nokia.com>
Reviewed-by: Mike Holmes <mike.holmes@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
Added recursive read-write lock. It allows threads to
read lock or write lock the rwlock multiple times without
deadlocking. Mixing of read and write lock operations is
not supported.
ODP version of recursive rwlock enables porting legacy
applications, which use these kind of locks.
Signed-off-by: Petri Savolainen <petri.savolainen@nokia.com>
Reviewed-by: Bill Fischofer <bill.fischofer@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
Conflicts:
platform/linux-generic/Makefile.am
|
|
Applications can use recursive spinlocks to avoid deadlock from
single thread acquiring the same lock multiple times. Recursive
locks are used in legacy applications. ODP version of recursive
spinlock enable porting of those applications.
Signed-off-by: Petri Savolainen <petri.savolainen@nokia.com>
Reviewed-by: Bill Fischofer <bill.fischofer@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
Conflicts:
platform/linux-generic/Makefile.am
|
|
Signed-off-by: Nicolas Morey-Chaisemartin <nmorey@kalray.eu>
Reviewed-by: Petri Savolainen <petri.savolainen@nokia.com>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
Added maximum thread count define to enable static memory allocation.
The API call returns dynamic maximum. Static and dynamic may be different
application is build for binary compatibility.
Signed-off-by: Petri Savolainen <petri.savolainen@nokia.com>
Reviewed-by: Bill Fischofer <bill.fischofer@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
Signed-off-by: Ivan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Reviewed-by: Bill Fischofer <bill.fischofer@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
Add API function for printing implementation specific pktio
debug information.
Signed-off-by: Matias Elo <matias.elo@nokia.com>
Reviewed-by: Petri Savolainen <petri.savolainen@nokia.com>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
Conflicts:
platform/linux-generic/odp_packet_io.c
|
|
If setting up the veth interfaces used for validation testing of socket
based pktios fails (e.g. CONFIG_VETH missing from the kernel), report
the test as skipped. Previously the test would be run anyway, resulting
in a lot of errors which aren't particularly easy to interpret.
Signed-off-by: Stuart Haslam <stuart.haslam@linaro.org>
Reviewed-by: Christophe Milard <christophe.milard@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
The netmap header file has some dubious casts that generate errors when
-Wcast-qual is enabled. Include it as a system header so that the
compiler ignores those errors without having to ignore errors in our own
source files.
Signed-off-by: Stuart Haslam <stuart.haslam@linaro.org>
Reviewed-by: Bill Fischofer <bill.fischofer@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
Timer API requires event types instead of buffer types.
This patch corrects it.
Signed-off-by: Ivan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Reviewed-by: Ola Liljedahl <ola.liljedahl@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
pktio/tap.c:267:21: error: comparison between signed and
unsigned integer expressions [-Werror=sign-compare]
} else if (retval != pkt_len) {
Cc: Ilya Maximets <i.maximets@samsung.com>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
Reviewed-by: Ilya Maximets <i.maximets@samsung.com>
Reviewed-by: Bill Fischofer <bill.fischofer@linaro.org>
|
|
Signed-off-by: Ilya Maximets <i.maximets@samsung.com>
Reviewed-and-tested-by: Stuart Haslam <stuart.haslam@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
Creates a new pktio type that allows for creating and
sending/receiving packets through TAP interface.
It is intended for use as a simple conventional communication
method between applications that use kernel network stack
(ping, ssh, iperf, etc.) and ODP applications for the purpose
of functional testing and can be used as it is with some
of the existing example applications.
To use this interface the name passed to odp_pktio_open() must
begin with "tap:" and be in the format:
tap:iface
iface the name of TAP device to be created.
TUN/TAP kernel module should be loaded to use this pktio.
There should be no device named 'iface' in the system.
The total length of the 'iface' is limited by IF_NAMESIZE.
Signed-off-by: Ilya Maximets <i.maximets@samsung.com>
Reviewed-and-tested-by: Stuart Haslam <stuart.haslam@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
This corrects bug https://bugs.linaro.org/show_bug.cgi?id=1921
Signed-off-by: Bill Fischofer <bill.fischofer@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
Add the new config tests to the TESTS list so that they get run via
"make check".
Signed-off-by: Stuart Haslam <stuart.haslam@linaro.org>
Reviewed-and-Tested-by: Anders Roxell <anders.roxell@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
Also debug output enhanced.
Signed-off-by: Ilya Maximets <i.maximets@samsung.com>
Reviewed-by: Stuart Haslam <stuart.haslam@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
Signed-off-by: Nicolas Morey-Chaisemartin <nmorey@kalray.eu>
Reviewed-by: Stuart Haslam <stuart.haslam@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
stopped
Signed-off-by: Nicolas Morey-Chaisemartin <nmorey@kalray.eu>
Reviewed-by: Stuart Haslam <stuart.haslam@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
sending
Signed-off-by: Nicolas Morey-Chaisemartin <nmorey@kalray.eu>
Reviewed-by: Stuart Haslam <stuart.haslam@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
reorder_prev is set to the address of the pointer origin_qe->s.reorder_head,
which is wrong. If the linked list was empty, that won't be corrected, and
reorder_prev->next points to the adjacent queue entry's status field. If that
entry is used, that queue's metadata will be corrupted.
This was found by running the chaos scheduler test with ODP-DPDK.
Signed-off-by: Zoltan Kiss <zoltan.kiss@linaro.org>
Reviewed-by: Bill Fischofer <bill.fischofer@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
|
|
If tick is late then application should be warned about this.
It means that actual timer resolution is worse than expected.
The default timer resolution is set 10ms, that is equal to jiffy
on most systems. The default resolution is set bearing in mind
that on a CPU runs maximum two threads that ideally fits in 10ms.
But user can change it to be smaller, in case if CPU0 is isolated
and it handles only the timer ticks. This patch helps user to set
correct timer resolution.
Signed-off-by: Ivan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Reviewed-by: Ola Liljedahl <ola.liljedahl@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
By default, all control threads execute on CPU 0 and
odp_cpumask_default_control() returns this CPU. For reasons of
performance, control and worker CPU's shouldn't overlap but for some
scenarios it can be desirable to utilize all CPU's for worker threads.
Thus we leave the decision of CPU allocation to the user but report
when a CPU is used for both control and worker threads.
Signed-off-by: Ivan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Reviewed-by: Ola Liljedahl <ola.liljedahl@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
|
|
Add a run-test script that replicates some of the "make check"
functionality without depending on a whole load of Makefile gubbins.
This makes it easier to perform on-target testing following a
cross-build (as we want to do in CI) and gives a simple way to verify
that all the required components do actually get installed correctly,
using "make installcheck".
./configure --enable-test-vald --with-testdir
DESTDIR=$(pwd)/test-install make install installcheck
Signed-off-by: Stuart Haslam <stuart.haslam@linaro.org>
Reviewed-by: Anders Roxell <anders.roxell@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
Previously if a build had been configured with validation tests enabled
(--enable-test-vald) a subsequent "make install" would install the test
binaries and scripts into /usr/local/bin without any way of separating
them from the rest of the installed artefacts. This patch changes the
default behaviour to not install test binaries or scripts and adds a
configure parameter (--with-testdir) to explicitly enable installation.
These changes are required in order to be able to package the tests
separately (e.g. in OpenEmbedded) and also to enable post-install
testing.
Signed-off-by: Stuart Haslam <stuart.haslam@linaro.org>
Reviewed-by: Anders Roxell <anders.roxell@linaro.org>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
Signed-off-by: Nicolas Morey-Chaisemartin <nmorey@kalray.eu>
Reviewed-by: Petri Savolainen <petri.savolainen@nokia.com>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
Signed-off-by: Nicolas Morey-Chaisemartin <nmorey@kalray.eu>
Reviewed-by: Petri Savolainen <petri.savolainen@nokia.com>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
For the moment just a simple wrapper around buffer_alloc to be used by
odp_buffer_alloc_multi and odp_packet_alloc_multi
Signed-off-by: Nicolas Morey-Chaisemartin <nmorey@kalray.eu>
Reviewed-by: Petri Savolainen <petri.savolainen@nokia.com>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|
|
Signed-off-by: Nicolas Morey-Chaisemartin <nmorey@kalray.eu>
Reviewed-by: Petri Savolainen <petri.savolainen@nokia.com>
Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org>
|