[lng-odp] howto build & run validation suite on any implementation?

I'd like to run the ODP validation suite ( test/common_plat/validation/api ) and examples against an implementation, but noticed it can only be built against an implementation that exists in the top-level platform/ directory of the main repository and that there seems to be no portable (universal)

[lng-odp] [PATCH] helper: remove unused variable

Signed-off-by: Mike Holmes --- helper/linux.c | 1 - 1 file changed, 1 deletion(-) diff --git a/helper/linux.c b/helper/linux.c index bb4b22c..7bd0b07 100644 --- a/helper/linux.c +++ b/helper/linux.c @@ -295,7 +295,6 @@ static int odph_linux_process_create(odph_odpthread_t *thread_tbl,

[lng-odp] [Bug 2478] scheduler.c bctx->sequence == qctx->lock_sequence[ndx]

https://bugs.linaro.org/show_bug.cgi?id=2478 Mike Holmes changed: What|Removed |Added Resolution|NON REPRODUCIBLE|--- Status|RESOLVED

Re: [lng-odp] [API-NEXT PATCH v4 0/5] new ordered queue implementation

For this series: Reviewed-and-tested-by: Bill Fischofer On Fri, Dec 2, 2016 at 4:56 AM, Matias Elo wrote: > V4: > - Rebased > - Added more static asserts to ordered lock functions (Bill) > > V3: > - Removed old SCHEDULE_ORDERED_LOCKS_PER_QUEUE define (Bill) > - Replaced error checks with assert

Re: [lng-odp] [PATCH] enable shared library for non-abi compat mode

On 2 December 2016 at 15:30, Maxim Uvarov wrote: > Signed-off-by: Maxim Uvarov > --- > configure.ac | 1 - > 1 file changed, 1 deletion(-) > > diff --git a/configure.ac b/configure.ac > index 3e89b0a..57edd33 100644 > --- a/configure.ac > +++ b/configure.ac > @@ -253,7 +253,6 @@ AC_ARG_ENABLE([a

Re: [lng-odp] [API-NEXT 1/2] api: init: add initializer function for odp_platform_init_t

On 12/02/16 17:55, Nicolas Morey-Chaisemartin wrote: > > Le 12/02/2016 à 03:42 PM, Maxim Uvarov a écrit : >> if you provide NULL is has to be initialized to default, do not see >> reason for separate api for that. >> >> Maxim. > Because if you want to provide only 1 settings among many, you needs

Re: [lng-odp] [API-NEXT PATCH v3 4/5] linux-gen: sched: new ordered queue implementation

On Fri, Dec 2, 2016 at 7:57 AM, Savolainen, Petri (Nokia - FI/Espoo) wrote: > > >> -Original Message- >> From: Bill Fischofer [mailto:bill.fischo...@linaro.org] >> Sent: Friday, December 02, 2016 3:12 PM >> To: Savolainen, Petri (Nokia - FI/Espoo) > labs.com> >> Cc: Elo, Matias (Nokia - FI

Re: [lng-odp] [API-NEXT 1/2] api: init: add initializer function for odp_platform_init_t

Le 12/02/2016 à 03:42 PM, Maxim Uvarov a écrit : > if you provide NULL is has to be initialized to default, do not see > reason for separate api for that. > > Maxim. Because if you want to provide only 1 settings among many, you needs to make sure all other are set to defaults Nicolas > > On 12

Re: [lng-odp] [PATCH] configure.ac update version numbers

Merged, Maxim. On 12/02/16 17:31, Mike Holmes wrote: > > > On 1 December 2016 at 16:05, Maxim Uvarov > wrote: > > Update numbers for .so and add short description. > > Signed-off-by: Maxim Uvarov > > > > Reviewed-

Re: [lng-odp] [API-NEXT 1/2] api: init: add initializer function for odp_platform_init_t

if you provide NULL is has to be initialized to default, do not see reason for separate api for that. Maxim. On 12/02/16 13:25, Nicolas Morey-Chaisemartin wrote: > Signed-off-by: Nicolas Morey-Chaisemartin > --- > include/odp/api/spec/init.h | 8 > platform/linux-generic/odp_init

Re: [lng-odp] [PATCH] configure.ac update version numbers

On 1 December 2016 at 16:05, Maxim Uvarov wrote: > Update numbers for .so and add short description. > > Signed-off-by: Maxim Uvarov > Reviewed-by: Mike Holmes Lib diff tool output http://people.linaro.org/~mike.holmes/diff-abi/ I think in the next point release we need to have yet another d

Re: [lng-odp] [PATCHv10 1/3] configure.ac: disable shared library for non abi compat mode

ok I sent patch to enable it. Please review. If it can be applied after current release let me know. To not wait again 24 hows for that. Maxim. On 12/02/16 17:21, Savolainen, Petri (Nokia - FI/Espoo) wrote: > > >> -Original Message- >> From: Maxim Uvarov [mailto:maxim.uva...@linaro.or

[lng-odp] [PATCH] enable shared library for non-abi compat mode

Signed-off-by: Maxim Uvarov --- configure.ac | 1 - 1 file changed, 1 deletion(-) diff --git a/configure.ac b/configure.ac index 3e89b0a..57edd33 100644 --- a/configure.ac +++ b/configure.ac @@ -253,7 +253,6 @@ AC_ARG_ENABLE([abi-compat], [if test "x$enableval" = "xno"; then ODP_ABI

Re: [lng-odp] [PATCHv10 1/3] configure.ac: disable shared library for non abi compat mode

> -Original Message- > From: Maxim Uvarov [mailto:maxim.uva...@linaro.org] > Sent: Friday, December 02, 2016 3:42 PM > To: Savolainen, Petri (Nokia - FI/Espoo) labs.com>; Mike Holmes > Cc: lng-odp > Subject: Re: [lng-odp] [PATCHv10 1/3] configure.ac: disable shared library > for non ab

Re: [lng-odp] [MONARCH PATCH 0/5] Make lidodp/libodp helper name platform specific

Le 12/02/2016 à 02:52 PM, Mike Holmes a écrit : > > > On 2 December 2016 at 08:50, Nicolas Morey-Chaisemartin > wrote: > > > > Le 12/02/2016 à 02:45 PM, Maxim Uvarov a écrit : > > On 12/02/16 16:34, Nicolas Morey-Chaisemartin wrote: > >> > >> Le 12/02/201

Re: [lng-odp] [API-NEXT PATCH v3 4/5] linux-gen: sched: new ordered queue implementation

> -Original Message- > From: Bill Fischofer [mailto:bill.fischo...@linaro.org] > Sent: Friday, December 02, 2016 3:12 PM > To: Savolainen, Petri (Nokia - FI/Espoo) labs.com> > Cc: Elo, Matias (Nokia - FI/Espoo) ; lng- > odp-forward > Subject: Re: [lng-odp] [API-NEXT PATCH v3 4/5] linux-

Re: [lng-odp] [MONARCH PATCH 0/5] Make lidodp/libodp helper name platform specific

On 2 December 2016 at 08:50, Nicolas Morey-Chaisemartin wrote: > > > Le 12/02/2016 à 02:45 PM, Maxim Uvarov a écrit : > > On 12/02/16 16:34, Nicolas Morey-Chaisemartin wrote: > >> > >> Le 12/02/2016 à 02:30 PM, Maxim Uvarov a écrit : > >>> That needs to go to master first, than if needed back-por

Re: [lng-odp] [MONARCH PATCH 0/5] Make lidodp/libodp helper name platform specific

Le 12/02/2016 à 02:45 PM, Maxim Uvarov a écrit : > On 12/02/16 16:34, Nicolas Morey-Chaisemartin wrote: >> >> Le 12/02/2016 à 02:30 PM, Maxim Uvarov a écrit : >>> That needs to go to master first, than if needed back-ported to Monarch. >> Ok. >> >>> Nicolas, if you name library differently how wi

Re: [lng-odp] [MONARCH PATCH 0/5] Make lidodp/libodp helper name platform specific

On 12/02/16 16:34, Nicolas Morey-Chaisemartin wrote: > > > Le 12/02/2016 à 02:30 PM, Maxim Uvarov a écrit : >> That needs to go to master first, than if needed back-ported to Monarch. > > Ok. > >> >> Nicolas, if you name library differently how will application know how >> to link with it? > >

Re: [lng-odp] [PATCHv10 1/3] configure.ac: disable shared library for non abi compat mode

On 12/02/16 11:29, Savolainen, Petri (Nokia - FI/Espoo) wrote: > What is "broken logic" with non-ABI compatible shared lib? configure.ac has ABI-compat enabled and at the same time --enable-abi-compat option. I.e. enable what is already enabled. So change is to keep it enabled by default and chang

Re: [lng-odp] [MONARCH PATCH 0/5] Make lidodp/libodp helper name platform specific

Le 12/02/2016 à 02:30 PM, Maxim Uvarov a écrit : > That needs to go to master first, than if needed back-ported to Monarch. Ok. > > Nicolas, if you name library differently how will application know how > to link with it? 1) We provide a template makefile for all ODP apps. 2) We don't have ABI

Re: [lng-odp] [MONARCH PATCH 0/5] Make lidodp/libodp helper name platform specific

That needs to go to master first, than if needed back-ported to Monarch. Nicolas, if you name library differently how will application know how to link with it? I think that we can stay with fixed name for ABI compat mode and let platforms to change it in non ABI compat mode. Assuming that generi

Re: [lng-odp] [API-NEXT PATCH v3 4/5] linux-gen: sched: new ordered queue implementation

On Fri, Dec 2, 2016 at 4:51 AM, Savolainen, Petri (Nokia - FI/Espoo) wrote: > >> >> This code still has the following in do_schedule(): >> >> ordered = sched_cb_queue_is_ordered(qi); >> >> /* Do not cache ordered events locally to improve >> * parallelism. Ordered context can only be released >> *

[lng-odp] [MONARCH PATCH 5/5] example: use ODP_LIB_NAME to generate libodp name

Signed-off-by: Nicolas Morey-Chaisemartin --- example/Makefile.inc | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/example/Makefile.inc b/example/Makefile.inc index 70ba2c0..b14cd69 100644 --- a/example/Makefile.inc +++ b/example/Makefile.inc @@ -1,6 +1,6 @@ include $(top_src

[lng-odp] [MONARCH PATCH 4/5] helper: use ODP_LIB_NAME to generate libodp name

Signed-off-by: Nicolas Morey-Chaisemartin --- helper/test/Makefile.am | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/helper/test/Makefile.am b/helper/test/Makefile.am index 545db73..8073283 100644 --- a/helper/test/Makefile.am +++ b/helper/test/Makefile.am @@ -28,10 +28

[lng-odp] [MONARCH PATCH 3/5] test: use ODP_LIB_NAME to generate libodp name

Signed-off-by: Nicolas Morey-Chaisemartin --- test/Makefile.inc| 2 +- test/validation/Makefile.inc | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/test/Makefile.inc b/test/Makefile.inc index 5def923..bb1f40f 100644 --- a/test/Makefile.inc +++ b/test/Makefile.in

[lng-odp] [MONARCH PATCH 2/5] platform: move odp-linux specific rules to linux-generic

Signed-off-by: Nicolas Morey-Chaisemartin --- configure.ac | 2 -- platform/Makefile.inc | 4 platform/linux-generic/Makefile.am | 6 ++ platform/linux-generic/m4/configure.m4 | 4 +++- 4 files changed, 9 insertions(+), 7 deletions(-) diff

[lng-odp] [MONARCH PATCH 1/5] m4: platform exports ODP_LIB_NAME

This is the suffix to be used to compute the name of ODP and ODP helper libraries. Signed-off-by: Nicolas Morey-Chaisemartin --- configure.ac | 1 + platform/linux-generic/m4/configure.m4 | 2 ++ 2 files changed, 3 insertions(+) diff --git a/configure.ac b/configure.a

[lng-odp] [MONARCH PATCH 0/5] Make lidodp/libodp helper name platform specific

The lib suffix "linux" (libodp-linux, libodphelper-linux) is harcoded pretty much everywhere in the build system. The goal of this series is to allow each platform to define its library suffix so they don't have to touch the generic files but simply the platform specific files. Nicolas Morey-Ch

[lng-odp] [API-NEXT PATCH v4 2/5] linux-gen: sched: remove old ordered queue implementation

Remove old ordered queue code. Replaced temporarily by atomic handling. Signed-off-by: Matias Elo --- platform/linux-generic/Makefile.am | 3 - .../linux-generic/include/odp_buffer_internal.h| 7 - .../linux-generic/include/odp_packet_io_queue.h| 5 +- .../linux-gen

[lng-odp] [API-NEXT PATCH v4 4/5] linux-gen: sched: new ordered queue implementation

Add new implementation for ordered queues. Compared to the old implementation this is much simpler and improves performance ~1-4x depending on the test case. The implementation is based on an atomic ordered context, which only a single thread may possess at a time. Only the thread owning the atomi

[lng-odp] [API-NEXT PATCH v4 3/5] linux-gen: sched: add internal API for max number of ordered locks per queue

The number of supported ordered locks may vary between the scheduler implementations. Add an internal scheduler API call for fetching the maximum value from currently active scheduler. Add an internal definition CONFIG_QUEUE_MAX_ORD_LOCKS for the scheduler independent maximum value. Signed-off-by

[lng-odp] [API-NEXT PATCH v4 5/5] linux-gen: sched: new ordered lock implementation

Implement ordered locks using per lock atomic counters. The counter values are compared to the queue’s atomic context to guarantee ordered locking. Compared to the previous implementation this enables parallel processing of ordered events outside of the lock context. Signed-off-by: Matias Elo ---

[lng-odp] [API-NEXT PATCH v4 1/5] linux-gen: sched: add internal APIs for locking/unlocking ordered processing

The internal ordered processing locking functions can be more streamlined compared to the public API functions. Signed-off-by: Matias Elo --- platform/linux-generic/include/odp_schedule_if.h | 4 platform/linux-generic/odp_schedule.c| 12 +++- platform/linux-generic/odp

[lng-odp] [API-NEXT PATCH v4 0/5] new ordered queue implementation

V4: - Rebased - Added more static asserts to ordered lock functions (Bill) V3: - Removed old SCHEDULE_ORDERED_LOCKS_PER_QUEUE define (Bill) - Replaced error checks with asserts in ordered lock/unlock (Bill) V2: - Support for multiple ordered locks (Bill) - New ordered lock implementation Add new

Re: [lng-odp] [API-NEXT PATCH v3 4/5] linux-gen: sched: new ordered queue implementation

> > This code still has the following in do_schedule(): > > ordered = sched_cb_queue_is_ordered(qi); > > /* Do not cache ordered events locally to improve > * parallelism. Ordered context can only be released > * when the local cache is empty. */ > if (ordered && max_num < MAX_DEQ) > max_de

[lng-odp] [API-NEXT 2/2] validation: init: add test for init with platform parameters

Signed-off-by: Nicolas Morey-Chaisemartin --- test/common_plat/validation/api/init/init.c | 17 + test/common_plat/validation/api/init/init.h | 1 + 2 files changed, 18 insertions(+) diff --git a/test/common_plat/validation/api/init/init.c b/test/common_plat/validation/api/init

[lng-odp] [API-NEXT 1/2] api: init: add initializer function for odp_platform_init_t

Signed-off-by: Nicolas Morey-Chaisemartin --- include/odp/api/spec/init.h | 8 platform/linux-generic/odp_init.c | 5 + 2 files changed, 13 insertions(+) diff --git a/include/odp/api/spec/init.h b/include/odp/api/spec/init.h index 154cdf8..0600691 100644 --- a/include/odp/api/

Re: [lng-odp] [API-NEXT PATCH v3 5/5] linux-gen: sched: new ordered lock implementation

The test above should be open code since it's a validity check on the call. The current code treats odp_schedule_order_lock/unlock as no-ops if we aren't running in an ordered context as that permits queue types to be moved from say ordered to atomic without having to change the application. The

Re: [lng-odp] [PATCH 1/4] configure.ac: do not set linux-generic as default platform

Just realized I didn't update the help string. One of a side effect of this series is there is NO default platform anymore. I'll wait for some feedback and repost a v2 if needed Le 12/02/2016 à 10:49 AM, Nicolas Morey-Chaisemartin a écrit : > This prepares for the second patch and add support for

[lng-odp] [PATCH 4/4] configure.ac: move linux-generic conditionals to platform side

Signed-off-by: Nicolas Morey-Chaisemartin --- configure.ac | 4 platform/linux-generic/m4/conditionals.m4 | 4 2 files changed, 4 insertions(+), 4 deletions(-) create mode 100644 platform/linux-generic/m4/conditionals.m4 diff --git a/configure.ac b/configu

[lng-odp] [PATCH 2/4] bootstrap: auto generate platform list

This list all the available platforms looking for platform/*/m4/configure.m4 and generate the appropriate m4 file for the configure.ac to include Signed-off-by: Nicolas Morey-Chaisemartin --- bootstrap| 25 + configure.ac | 11 +-- 2 files changed, 26 insert

[lng-odp] [PATCH 3/4] bootstrap: auto-include platform specific conditional definitions

Conditionals must always be defined if reference in a Makefile.am even if the Makefile does not apply to the platform being built. As long as all the conditional are defined with default values in platform//m4/conditionals.m4, bootstrap will automatically include them so there won't be any issue w

[lng-odp] [PATCH 1/4] configure.ac: do not set linux-generic as default platform

This prepares for the second patch and add support for repo which do not have a linux-generic platform Signed-off-by: Nicolas Morey-Chaisemartin --- configure.ac | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/configure.ac b/configure.ac index be5a292..86173f4 100644 --- a/c

[lng-odp] [PATCH 0/4] Make configure.ac generic

The goal of this patch series is to auto include all avalaible platform and not tie the configure.ac to linux-generic only. It basically means that a platform could simply be a submodule clone into plat/ and be automatically included and compile with the appropriate --with-platform= option, with

Re: [lng-odp] cannot create multiple threads for traffic management

Hi Barry, Sorry to call you surname. I have changed the number to 4 cores. Please have a review for http://patches.opendataplane.org/patch/7515/ Thanks, Forrest On 30 November 2016 at 15:47, Forrest Shi wrote: > Hi Spinney, > > I noticed you add a tm group to the traffic management and create

[lng-odp] [RFC v2] example: traffic_mgmt: enhancement with multiple pktios

From: Xuelin Shi This patch made the following changes: - try to create one tm background thread for each pktio - receiving packets from multiple pktios other than generating packets - identifying TM user by ip with individual class of service - print the user service while starting the p

[lng-odp] [PATCH] linux-generic: move tm system barrier to tm group

From: Xuelin Shi since tm thread is handling tm group, move the thread based barrier to tm group. otherwise, packet cannot get into the second tm system in the same group. Signed-off-by: Xuelin Shi --- platform/linux-generic/include/odp_traffic_mngr_internal.h | 3 ++- platform/linux-generic/

[lng-odp] [MONARCH PATCH] doc: remove VPATH override

It breaks out-of-tree build of documentation as .adoc file are not in the VPATH and default rules cannot be applied to them. Signed-off-by: Nicolas Morey-Chaisemartin --- doc/Makefile.inc | 2 -- 1 file changed, 2 deletions(-) diff --git a/doc/Makefile.inc b/doc/Makefile.inc index 643b1d4..170

[lng-odp] [PATCH] linux-generic: only enable pktout when egress kind is pktio

From: Xuelin Shi Signed-off-by: Xuelin Shi --- platform/linux-generic/odp_traffic_mngr.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/platform/linux-generic/odp_traffic_mngr.c b/platform/linux-generic/odp_traffic_mngr.c index ffb149b..a1f990f 100644 --- a/platform/l

[lng-odp] [API-NEXT PATCH] linux-gen: _ishm: unlinking files asap for cleaner termination

_ishm now unlinks the created files as soon as possible, hence reducing the chance to see left-overs, if ODP terminates abnormaly. This does not provide 100% guaranty: if we are unlucky enough, ODP may be killed between open() and unlink(). Also this method excludes exported files (flag _ODP_ISHM_E

Re: [lng-odp] [PATCHv10 1/3] configure.ac: disable shared library for non abi compat mode

> -Original Message- > From: lng-odp [mailto:lng-odp-boun...@lists.linaro.org] On Behalf Of Maxim > Uvarov > Sent: Thursday, December 01, 2016 9:50 PM > To: Mike Holmes > Cc: lng-odp > Subject: Re: [lng-odp] [PATCHv10 1/3] configure.ac: disable shared library > for non abi compat mode >