Re: [lng-odp] [RFC v2 0/4] Move the definition of odp syncronizers abstract types to platform

On further analysis the make distcheck errors I reported for the parts of this seem to be artifacts of check-odp. I can't make them happen manually so I'm not sure what the issue might be. Anders: Can you take a look at this and advise? In any event, the patches apply cleanly and all tests pass,

Re: [lng-odp] [RFC v2 4/4] barrier: allow platform to override odp_barrier_t

This part fails make distcheck with error: CC odp_barrier.lo In file included from ../../../platform/linux-generic/odp_barrier.c:7:0: ../../../platform/linux-generic/include/odp/barrier.h:22:36: fatal error: odp/plat/barrier_types.h: No such file or directory #include

Re: [lng-odp] [RFC v2 3/4] ticketlock: allow platform to override odp_ticketlock_t

This part also fails make distcheck with error: CC odp_buffer.lo In file included from ../../../platform/linux-generic/include/odp_pool_internal.h:65:0, from ../../../platform/linux-generic/odp_buffer.c:8: ../../../platform/linux-generic/include/odp/ticketlock.h:20:39: fat

Re: [lng-odp] [RFC v2 2/4] rwlock: allow platform to override odp_rwlock_t

This part also fails make distcheck with error: CC odp_classification.lo In file included from ../../../platform/linux-generic/include/odp_packet_io_internal.h:21:0, from ../../../platform/linux-generic/odp_classification.c:15: ../../../platform/linux-generic/include/odp/s

Re: [lng-odp] [RFC v2 1/4] spinlock: allow platform to override odp_spinlock_t

This part fails make distcheck with error: In file included from ../../../platform/linux-generic/include/odp_packet_io_internal.h:21:0, from ../../../platform/linux-generic/odp_classification.c:15: ../../../platform/linux-generic/include/odp/spinlock.h:20:37: fatal error: odp/plat

Re: [lng-odp] [PATCHv2 0/3] Fixes for https://bugs.linaro.org/show_bug.cgi?id=1284

For this series: Reviewed-and-tested-by: Bill Fischofer On Tue, Mar 17, 2015 at 4:53 AM, Ciprian Barbu wrote: > v2: > - fixed patch 3 > > Like in the previous RFC, the first patch is the fix. > The second and third patch represent a proposed fix for freeing the > ODP_QUEUE_TYPE_SCHED queues th

Re: [lng-odp] [PATCH] performance: odp_atomic: move atomic test from api_test to performance

These changes do not seem to be a simple move as logic is being changed. A better commit message needed to say what's really happening perhaps? On Tue, Mar 17, 2015 at 3:59 PM, Mike Holmes wrote: > Signed-off-by: Mike Holmes > --- > test/api_test/.gitignore | 1 - >

[lng-odp] [PATCH] performance: odp_atomic: move atomic test from api_test to performance

Signed-off-by: Mike Holmes --- test/api_test/.gitignore | 1 - test/api_test/Makefile.am | 6 +- test/api_test/odp_atomic_test.h| 51 test/api_test/odp_common.c | 2 +- test/performance/

[lng-odp] [PATCH] move odp_atomic from api_test to performance directory

The odp_atomic test can form the basis for testing the scalability of the atomic api. Move it so that the api_test dir may be deleted and the odp_atomic test can be run as part of test/performance Mike Holmes (1): performance: odp_atomic: move atomic test from api_test to performance test/api

Re: [lng-odp] [PATCH] doc: doxygen: add declaring variables

On Tue, Mar 17, 2015 at 1:57 PM, Mike Holmes wrote: > Signed-off-by: Mike Holmes > Reviewed-by: Bill Fischofer > --- > doc/api_guide_lines.dox | 15 +++ > 1 file changed, 15 insertions(+) > > diff --git a/doc/api_guide_lines.dox b/doc/api_guide_lines.dox > index 4cfe088..2bc63a1

[lng-odp] [PATCH] doc: doxygen: add declaring variables

Signed-off-by: Mike Holmes --- doc/api_guide_lines.dox | 15 +++ 1 file changed, 15 insertions(+) diff --git a/doc/api_guide_lines.dox b/doc/api_guide_lines.dox index 4cfe088..2bc63a1 100644 --- a/doc/api_guide_lines.dox +++ b/doc/api_guide_lines.dox @@ -149,6 +149,21 @@ When an inte

Re: [lng-odp] [PATCH] doc: move api guidelines to API doc

Merged, Maxim. On 03/17/15 15:51, Mike Holmes wrote: ping On 11 March 2015 at 14:43, Bill Fischofer > wrote: On Wed, Mar 11, 2015 at 1:24 PM, Mike Holmes mailto:mike.hol...@linaro.org>> wrote: This documentation was moved from the architect

[lng-odp] [PATCHv3] example: l2fwd print packets per second

Current print in l2fwd is not useful with slow links and also it's hard to say how fast does it work. Print pps and packets drops. Signed-off-by: Maxim Uvarov --- v3: alloc memory for counter in worker threads to avoid cache misses. This patch is independent of 'make check' script. Sending it

[lng-odp] UberConference Reminder

UberConference Reminder___ lng-odp mailing list lng-odp@lists.linaro.org http://lists.linaro.org/mailman/listinfo/lng-odp

Re: [lng-odp] ipc for odp

On Tue, Mar 17, 2015 at 1:44 PM, Maxim Uvarov wrote: > On 03/17/15 14:30, Ciprian Barbu wrote: >> >> On Fri, Mar 13, 2015 at 5:30 PM, Ola Liljedahl >> wrote: >>> >>> On 13 March 2015 at 13:35, Maxim Uvarov wrote: On 03/13/15 13:43, Ciprian Barbu wrote: > > On Fri, Mar 13, 2015

[lng-odp] [Bug 1365] New: Lock-up following sendmmsg failure

https://bugs.linaro.org/show_bug.cgi?id=1365 Bug ID: 1365 Summary: Lock-up following sendmmsg failure Product: OpenDataPlane Version: 1.0 Hardware: Other OS: Linux Status: UNCONFIRMED Severity: enhancement

Re: [lng-odp] [PATCH] doc: move api guidelines to API doc

ping On 11 March 2015 at 14:43, Bill Fischofer wrote: > > > On Wed, Mar 11, 2015 at 1:24 PM, Mike Holmes > wrote: > >> This documentation was moved from the architecture doc to this API doc. >> >> Signed-off-by: Mike Holmes >> > > Reviewed-by: Bill Fischofer > > >> --- >> doc/api_guide_lines

Re: [lng-odp] [PATCH] doc: move api guidelines to API doc

ping On 11 March 2015 at 14:43, Bill Fischofer wrote: > > > On Wed, Mar 11, 2015 at 1:24 PM, Mike Holmes > wrote: > >> This documentation was moved from the architecture doc to this API doc. >> >> Signed-off-by: Mike Holmes >> > > Reviewed-by: Bill Fischofer > > >> --- >> doc/api_guide_lines

Re: [lng-odp] ipc for odp

On 03/17/15 14:30, Ciprian Barbu wrote: On Fri, Mar 13, 2015 at 5:30 PM, Ola Liljedahl wrote: On 13 March 2015 at 13:35, Maxim Uvarov wrote: On 03/13/15 13:43, Ciprian Barbu wrote: On Fri, Mar 13, 2015 at 11:34 AM, Ola Liljedahl wrote: When I think about IPC, it is message passing between

Re: [lng-odp] ipc for odp

On Fri, Mar 13, 2015 at 5:30 PM, Ola Liljedahl wrote: > On 13 March 2015 at 13:35, Maxim Uvarov wrote: >> >> On 03/13/15 13:43, Ciprian Barbu wrote: >>> >>> On Fri, Mar 13, 2015 at 11:34 AM, Ola Liljedahl >>> wrote: When I think about IPC, it is message passing between different progra

Re: [lng-odp] [PATCH 1/2] validation: timer: use ODP_SHM_NULL

Merged, Maxim. On 03/06/15 20:30, Mike Holmes wrote: On 6 March 2015 at 09:30, Petri Savolainen > wrote: Use ODP_SHM_NULL to select default shm memory in odp_pool_create(), instead of ODP_SHM_INVALID. Signed-off-by: Petri Savolainen mailto:petr

Re: [lng-odp] [PATCHv2 2/2] tests: performance add l2fwd to run script

On Mon, Mar 16, 2015 at 06:56:19PM +0300, Maxim Uvarov wrote: > On 03/16/15 14:09, Stuart Haslam wrote: > >On Fri, Mar 13, 2015 at 06:03:47PM +0300, Maxim Uvarov wrote: > >>Move common pktio linux-generic functions to setup virtual net to > >>linux-generic folder and add running l2fwd test for make

Re: [lng-odp] [PATCH] linux-generic: strongtypes: add dummy struct var to keep c++ happy

Merged, Maxim. On 03/13/15 22:29, Mike Holmes wrote: On 12 March 2015 at 17:36, Bill Fischofer > wrote: C++ doesn't like null structs so add a dummy variable to make it happy. Note that we only use these types as pseudo-pointers for strong typing

[lng-odp] [PATCHv2 1/3] validation: pktio: add support for VPATH builds

Running "make check" in a VPATH build fails as the odp_pktio_run script incorrectly assumes that the odp_pktio binary is in the same directory. Resolve this by passing the path to the test binary to the script via the TESTS_ENVIRONMENT variable. Signed-off-by: Stuart Haslam --- test/validation/M

[lng-odp] [PATCHv2 3/3] performance: scheduling: add support for VPATH builds

Pass the test executable path to the test script to avoid the assumption that they're both in the same directory. Signed-off-by: Stuart Haslam --- Change in v2; correct name of executable test/performance/Makefile.am| 2 ++ test/performance/odp_scheduling_run | 3 ++- 2 files changed, 4

[lng-odp] [PATCHv2 2/3] performance: scheduling: catch exit codes

odp_scheduling_run always exits with code 0 so "make check" will never report failures. Signed-off-by: Stuart Haslam --- Change in v2; correct name of executable test/performance/odp_scheduling_run | 6 +- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/test/performance/odp_sch

Re: [lng-odp] [PATCHv2] validation: timer: added cancel test

Merged, Maxim. On 03/17/15 04:38, Bill Fischofer wrote: For the additions: Reviewed-by: Bill Fischofer > On Mon, Mar 16, 2015 at 8:55 AM, Ola Liljedahl mailto:ola.liljed...@linaro.org>> wrote: On 12 March 2015 at 17:03, Ola Liljedahl mailto:ola.liljed.

Re: [lng-odp] [PATCH 2/3] performance: scheduling: catch exit codes

On Tue, Mar 17, 2015 at 11:44:10AM +0200, Ciprian Barbu wrote: > On Mon, Mar 16, 2015 at 1:48 PM, Stuart Haslam > wrote: > > odp_scheduling_run always exits with code 0 so "make check" will never > > report failures. > > > > Signed-off-by: Stuart Haslam > > --- > > test/performance/odp_scheduli

Re: [lng-odp] [PATCH] linux-generic: queue: fix potential queue leak

Merged, Maxim. On 03/17/15 04:40, Bill Fischofer wrote: On Mon, Mar 16, 2015 at 8:58 AM, Stuart Haslam mailto:stuart.has...@linaro.org>> wrote: If schedule buffer allocation fails during odp_queue_create(), a scheduled or pktin queue may be incorrectly left in an allocated state

[lng-odp] [RFC v2 2/4] rwlock: allow platform to override odp_rwlock_t

Signed-off-by: Jerin Jacob --- include/odp/api/rwlock.h | 11 + .../linux-generic/include/odp/plat/rwlock_types.h | 48 ++ platform/linux-generic/include/odp/rwlock.h| 2 + 3 files changed, 52 insertions(+), 9 deletions(-) create mode 1

[lng-odp] [RFC v2 3/4] ticketlock: allow platform to override odp_ticketlock_t

Signed-off-by: Jerin Jacob --- include/odp/api/ticketlock.h | 12 +- .../include/odp/plat/ticketlock_types.h| 46 ++ platform/linux-generic/include/odp/ticketlock.h| 2 + 3 files changed, 50 insertions(+), 10 deletions(-) create mode

[lng-odp] [RFC v2 4/4] barrier: allow platform to override odp_barrier_t

Signed-off-by: Jerin Jacob --- include/odp/api/barrier.h | 7 +--- platform/linux-generic/include/odp/barrier.h | 1 + .../linux-generic/include/odp/plat/barrier_types.h | 47 ++ 3 files changed, 49 insertions(+), 6 deletions(-) create mode 10

[lng-odp] [RFC v2 0/4] Move the definition of odp syncronizers abstract types to platform

Move the definition of odp syncronizers abstract types to platform from public headerfile. This will allow platform define odp syncronizers abstract type. Useful when native SDK has definition of the odp syncronizers and ODP implementation decides to reuses them. v1..v2 Corrected Doxygen documenta

[lng-odp] [RFC v2 1/4] spinlock: allow platform to override odp_spinlock_t

Signed-off-by: Jerin Jacob --- include/odp/api/spinlock.h | 10 + .../include/odp/plat/spinlock_types.h | 46 ++ platform/linux-generic/include/odp/spinlock.h | 2 + 3 files changed, 50 insertions(+), 8 deletions(-) create mode 1

Re: [lng-odp] [PATCHv2] validation: pktio: prevent race when using veth pair

Merged, Maxim. On 03/16/15 15:27, Stuart Haslam wrote: There's a potential race condition whereby the test case could start running before the virtual ethernet interfaces are fully brought up. So replace the the arbitrary delay with a check for the interface's operational state before kicking of

[lng-odp] [PATCHv2 0/3] Fixes for https://bugs.linaro.org/show_bug.cgi?id=1284

v2: - fixed patch 3 Like in the previous RFC, the first patch is the fix. The second and third patch represent a proposed fix for freeing the ODP_QUEUE_TYPE_SCHED queues that were not removed from the internal priority queues because the scheduler did not know they were empty. Ciprian Barbu (3):

[lng-odp] [PATCHv2 2/3] linux-generic: queue: add queue_is_destroyed internal call

Signed-off-by: Ciprian Barbu --- platform/linux-generic/include/odp_queue_internal.h | 9 + 1 file changed, 9 insertions(+) diff --git a/platform/linux-generic/include/odp_queue_internal.h b/platform/linux-generic/include/odp_queue_internal.h index 65aae14..9a0e94f 100644 --- a/platform

[lng-odp] [PATCHv2 1/3] validation: schedule: free queues and pool

With the implementation of termination APIs it is now necessarry to cleanup all allocated resources, queues and pool in this case. Fixes https://bugs.linaro.org/show_bug.cgi?id=1284 Signed-off-by: Ciprian Barbu --- test/validation/odp_schedule.c | 56 -- 1

[lng-odp] [PATCHv2 3/3] linux-generic: schedule: make sure SCHED queues get freed by the scheduler

ODP_QUEUE_TYPE_SCHED queues only get marked as destroyed, they need to be removed from the pri_queues of the linux-generic scheduler Signed-off-by: Ciprian Barbu --- v2: - removed #if 1 and trailing whitespaces platform/linux-generic/odp_schedule.c | 22 ++ 1 file changed, 2

Re: [lng-odp] [PATCH 3/3] linux-generic: schedule: make sure SCHED queues get freed by the scheduler

On Tue, Mar 17, 2015 at 3:43 AM, Bill Fischofer wrote: > Code should not have #if 0 or #if 1 in them. Leftover from debugging? Yup, sorry about that. > > On Mon, Mar 16, 2015 at 10:16 AM, Maxim Uvarov > wrote: >> >> On 03/16/15 17:51, Ciprian Barbu wrote: >>> >>> ODP_QUEUE_TYPE_SCHED queues on

Re: [lng-odp] [PATCH 2/3] performance: scheduling: catch exit codes

On Mon, Mar 16, 2015 at 1:48 PM, Stuart Haslam wrote: > odp_scheduling_run always exits with code 0 so "make check" will never > report failures. > > Signed-off-by: Stuart Haslam > --- > test/performance/odp_scheduling_run | 6 +- > 1 file changed, 5 insertions(+), 1 deletion(-) > > diff --g

Re: [lng-odp] [RFC 1/4] spinlock: allow platform to override odp_spinlock_t

All these changes are needed. Just copy the mechanism from other abstract typedefs we have already. See comments under. > -Original Message- > From: lng-odp-boun...@lists.linaro.org [mailto:lng-odp- > boun...@lists.linaro.org] On Behalf Of ext Jerin Jacob > Sent: Tuesday, March 17, 2015 9

[lng-odp] [RFC 3/4] ticketlock: allow platform to override odp_ticketlock_t

Signed-off-by: Jerin Jacob --- include/odp/api/ticketlock.h | 12 -- .../include/odp/plat/ticketlock_types.h| 43 ++ platform/linux-generic/include/odp/ticketlock.h| 2 + 3 files changed, 45 insertions(+), 12 deletions(-) create mode

[lng-odp] [RFC 4/4] barrier: allow platform to override odp_barrier_t

Signed-off-by: Jerin Jacob --- include/odp/api/barrier.h | 9 - platform/linux-generic/include/odp/barrier.h | 1 + .../linux-generic/include/odp/plat/barrier_types.h | 44 ++ 3 files changed, 45 insertions(+), 9 deletions(-) create mode 1

[lng-odp] [RFC 0/4] Move the definition of odp syncronizers abstract types to platform

Move the definition of odp syncronizers abstract types to platform from public headerfile. This will allow platform define odp syncronizers abstract type. Useful when native SDK has definition of the odp syncronizers and ODP implementation decides to reuses them. Jerin Jacob (4): spinlock: allow

[lng-odp] [RFC 2/4] rwlock: allow platform to override odp_rwlock_t

Signed-off-by: Jerin Jacob --- include/odp/api/rwlock.h | 12 -- .../linux-generic/include/odp/plat/rwlock_types.h | 45 ++ platform/linux-generic/include/odp/rwlock.h| 2 + 3 files changed, 47 insertions(+), 12 deletions(-) create mode

[lng-odp] [RFC 1/4] spinlock: allow platform to override odp_spinlock_t

Signed-off-by: Jerin Jacob --- include/odp/api/spinlock.h | 10 -- .../include/odp/plat/spinlock_types.h | 42 ++ platform/linux-generic/include/odp/spinlock.h | 2 ++ 3 files changed, 44 insertions(+), 10 deletions(-) create mod