On Thu, May 28, 2015 at 12:58:10PM +0200, Christophe Milard wrote:
Module pktio now gets its own directory and create its own lib
(currentely only containing its executable)
Startup scripting stuff is just moved to the pktio directory but remains
untouched
checkpatch complains about this
On Thu, May 28, 2015 at 12:58:09PM +0200, Christophe Milard wrote:
Preparing for the next patch where this file is moved and
check-odp would yell if these things were still there.
Signed-off-by: Christophe Milard christophe.mil...@linaro.org
Reviewed-by: Stuart Haslam
Looks good. But why default is 1 and not BATCH_LEN_MAX? Then more then
faster right?
Stuart, it's your test. Do you also want to review it?
Thank you,
Maxim.
On 06/01/15 19:44, Nicolas Morey-Chaisemartin wrote:
Signed-off-by: Nicolas Morey-Chaisemartin nmo...@kalray.eu
---
On Mon, Jun 01, 2015 at 06:44:10PM +0200, Nicolas Morey-Chaisemartin wrote:
Signed-off-by: Nicolas Morey-Chaisemartin nmo...@kalray.eu
---
test/performance/odp_pktio_perf.c | 72
+--
1 file changed, 54 insertions(+), 18 deletions(-)
diff --git
On Thu, May 28, 2015 at 12:58:11PM +0200, Christophe Milard wrote:
When tests will be ran from the platform side, they will use
platform agnostic tests from the validation side: i.e.
-the validation side must be build before the platform test side.
And the platform agnostic tests uses ODP.
On 06/02/2015 11:19 AM, Stuart Haslam wrote:
On Mon, Jun 01, 2015 at 06:44:10PM +0200, Nicolas Morey-Chaisemartin wrote:
Signed-off-by: Nicolas Morey-Chaisemartin nmo...@kalray.eu
---
test/performance/odp_pktio_perf.c | 72
+--
1 file changed, 54
On Thu, May 28, 2015 at 12:58:08PM +0200, Christophe Milard wrote:
Renaming of things which may be, one day, exported in a lib.
This renaming is important, as it creates consistency between test
symbols, which is needed if things get eventually exported in the lib.
Also, tests are often
On Thu, May 28, 2015 at 12:58:07PM +0200, Christophe Milard wrote:
In odp_cunit_common.c, a macro, called MODULE_HAS_OWN_MAIN is used to tell
whether to define a main or not.
If MODULE_HAS_OWN_MAIN is defined, odp_cunit_common.c does not
define any main, but offers odp_cunit_run(CU_SuiteInfo
Zoltan,
I am currently working on ThunderX port so I can offer an insight into one
of the implementations.
ThunderX has more server-like network adapter as opposed to Octeon or
QorIQ, so buffer management is done in software.
I think the problem with pool starvation affects mostly those kinds of
On 05/31/15 23:38, Bill Fischofer wrote:
Implement card https://cards.linaro.org/browse/LNG-925
reference for card should go after --, it's internal information only.
And here should some readable description why we do this for
people who see only git logs.
Thanks,
Maxim.
Signed-off-by:
Because I wanted to keep the default behavior the same as it is now.
Using the same value for both rx and tx batch will definitely give the same
performance but the results will be biased.
In a normal behavior you can't expect your packet to be sent and received in
the exact same burst size all
On 06/02/15 11:52, Nicolas Morey-Chaisemartin wrote:
Because I wanted to keep the default behavior the same as it is now.
Using the same value for both rx and tx batch will definitely give the same
performance but the results will be biased.
In a normal behavior you can't expect your packet to
On 2 June 2015 at 11:03, Maxim Uvarov maxim.uva...@linaro.org wrote:
Signed-off-by: Maxim Uvarov maxim.uva...@linaro.org
Reviewed-by: Mike Holmes mike.hol...@linaro.org
---
scripts/checkpatch.pl | 2 ++
1 file changed, 2 insertions(+)
diff --git a/scripts/checkpatch.pl
On 2 June 2015 at 07:22, Stuart Haslam stuart.has...@linaro.org wrote:
When a file is added, deleted or renamed checkpatch produces a warning
about updating the MAINTAINERS file. Remove the warning as we don't
have a MAINTAINERS file.
Signed-off-by: Stuart Haslam stuart.has...@linaro.org
UberConference Cancelled___
lng-odp mailing list
lng-odp@lists.linaro.org
https://lists.linaro.org/mailman/listinfo/lng-odp
I have a problem with this - I might have missed something.
So I have codespell installed:-
mike@mike-desktop:~/git/odp$ sudo apt-get install codespell
Reading package lists... Done
Building dependency tree
Reading state information... Done
codespell is already the newest version.
0 upgraded, 0
On Tue, Jun 02, 2015 at 11:46:34AM +0200, Nicolas Morey-Chaisemartin wrote:
On 06/02/2015 11:19 AM, Stuart Haslam wrote:
On Mon, Jun 01, 2015 at 06:44:10PM +0200, Nicolas Morey-Chaisemartin wrote:
Signed-off-by: Nicolas Morey-Chaisemartin nmo...@kalray.eu
---
On 06/02/15 12:46, Nicolas Morey-Chaisemartin wrote:
On 06/02/2015 11:19 AM, Stuart Haslam wrote:
On Mon, Jun 01, 2015 at 06:44:10PM +0200, Nicolas Morey-Chaisemartin wrote:
Signed-off-by: Nicolas Morey-Chaisemartin nmo...@kalray.eu
---
test/performance/odp_pktio_perf.c | 72
Obvious remove. Merged.
Maxim.
On 05/27/15 18:50, Bill Fischofer wrote:
Signed-off-by: Bill Fischofer bill.fischo...@linaro.org
---
include/odp/api/packet.h | 14 --
1 file changed, 14 deletions(-)
diff --git a/include/odp/api/packet.h b/include/odp/api/packet.h
index
When a file is added, deleted or renamed checkpatch produces a warning
about updating the MAINTAINERS file. Remove the warning as we don't
have a MAINTAINERS file.
Signed-off-by: Stuart Haslam stuart.has...@linaro.org
---
scripts/checkpatch.pl | 2 +-
1 file changed, 1 insertion(+), 1
On 2 June 2015 at 11:34, Maciej Czekaj m...@semihalf.com wrote:
Zoltan,
I am currently working on ThunderX port so I can offer an insight into one
of the implementations.
ThunderX has more server-like network adapter as opposed to Octeon or
QorIQ, so buffer management is done in software.
Merged,
Thanks,
Maxim.
On 05/28/15 12:06, Jerin Jacob wrote:
use of CU_ASSERT(from == queue) in 'test_schedule_pause_resume'
odp_schedule.c: In function 'test_schedule_pause_resume':
odp_schedule.c:573:3: error: 'from' may be used uninitialized
in this function [-Werror=maybe-uninitialized]
On Wed, May 20, 2015 at 12:31:22PM +0300, Maxim Uvarov wrote:
Some platforms like DPDK and Cavium bare metal can not
dynamically destroy and create pool with the same name.
If this is expected behaviour then it should be mentioned in the API
documentation as it's currently not. Presumably these
Thinking about using notification events instead of the messages
themselves...
If I use D-BUS or ZeroMQ, I get a socket descriptor. I don't want to
block on this descriptor, my application is likely blocking in
odp_schedule() or polling some queues. I could allocate an event (of
whatever type)
v2 submitted. Maxim: Checkpatch is still flagging this:
CHECK: Avoid CamelCase: PRIu64
#304: FILE: platform/linux-generic/odp_pool.c:580:
+ ODP_DBG( pool storageODP managed shm handle % PRIu64 \n,
Can we get an update for this as this usage is not under our control.
On Tue, Jun 2, 2015 at
https://bugs.linaro.org/show_bug.cgi?id=1609
Bug ID: 1609
Summary: helpers: odph_ipv4_udp_chksum is not tested
Product: OpenDataPlane
Version: 1.0.4
Hardware: Other
OS: Linux
Status: UNCONFIRMED
Severity:
On 06/02/15 17:47, Bill Fischofer wrote:
v2 submitted. Maxim: Checkpatch is still flagging this:
CHECK: Avoid CamelCase: PRIu64
#304: FILE: platform/linux-generic/odp_pool.c:580:
+ODP_DBG( pool storageODP managed shm handle % PRIu64 \n,
Can we get an update for this as this usage is not
packet i/o test can create 2 types of queues: scheduled and
polled. Do not do dequeue from scheduled queue.
https://bugs.linaro.org/show_bug.cgi?id=1383
Signed-off-by: Maxim Uvarov maxim.uva...@linaro.org
---
test/validation/odp_pktio.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
This patch removes the 2nd parameter to odp_pool_create() because it has two
major issues:
1. Applications have no way to know how big a shm area is required for a given
pool since that information is implementation specific. So there's no portable
means of using this parameter.
2. Some
On Sun, May 31, 2015 at 03:38:55PM -0500, Bill Fischofer wrote:
Implement card https://cards.linaro.org/browse/LNG-925
Signed-off-by: Bill Fischofer bill.fischo...@linaro.org
---
example/classifier/odp_classifier.c| 2 +-
example/generator/odp_generator.c
OK, will post a v2 to cover both Maxim's and Stuart's comments. Thanks.
On Tue, Jun 2, 2015 at 6:08 AM, Stuart Haslam stuart.has...@linaro.org
wrote:
On Sun, May 31, 2015 at 03:38:55PM -0500, Bill Fischofer wrote:
Implement card https://cards.linaro.org/browse/LNG-925
Signed-off-by: Bill
Hi Ola, all,
On 05/27/2015 10:08 AM, Ola Liljedahl wrote:
Here is my third attempt at a ODP API for MSGIO - asynchronous message passing
based IPC for a shared nothing architecture.
I understand the value of this, but I feel I need to ask the question:
does it really belong to ODP ? My
On 06/02/2015 11:58 AM, Maxim Uvarov wrote:
Btw, is odp_queue_deq faster than odp_queue_deq_multi for 1 packet? If not
why do we need odp_queue_deq then?
Maxim.
From what I can see on our platform
odp_schedule and odp_schedule_multi with num = 1 have the same performances.
It seems that
We need a test for it in ~/git/odp/helper/test, it is one of the last few
APIs without its own test.
On 2 June 2015 at 09:37, Maxim Uvarov maxim.uva...@linaro.org wrote:
ping. Does anybody also reviewed that patch?
Thanks,
Maxim.
On 05/19/15 13:49, Maxim Uvarov wrote:
From: Alexandru
ping. Does anybody also reviewed that patch?
Thanks,
Maxim.
On 05/19/15 13:49, Maxim Uvarov wrote:
From: Alexandru Badicioiu alexandru.badici...@linaro.org
Signed-off-by: Alexandru Badicioiu alexandru.badici...@linaro.org
Signed-off-by: Maxim Uvarov maxim.uva...@linaro.org
---
v2: fix csum
https://bugs.linaro.org/show_bug.cgi?id=1603
Mike Holmes mike.hol...@linaro.org changed:
What|Removed |Added
CC||mike.hol...@linaro.org
On 2 June 2015 at 08:53, Jerin Jacob jerin.ja...@caviumnetworks.com wrote:
On Tue, Jun 02, 2015 at 02:36:47PM +0200, Benoît Ganne wrote:
Hi Ola, all,
On 05/27/2015 10:08 AM, Ola Liljedahl wrote:
Here is my third attempt at a ODP API for MSGIO - asynchronous message
passing
based IPC
Merged and created bug to add note:
https://bugs.linaro.org/show_bug.cgi?id=1608
Thanks,
Maxim.
On 06/02/15 14:10, Stuart Haslam wrote:
On Wed, May 20, 2015 at 12:31:22PM +0300, Maxim Uvarov wrote:
Some platforms like DPDK and Cavium bare metal can not
dynamically destroy and create pool with
On Tue, Jun 02, 2015 at 02:36:47PM +0200, Benoît Ganne wrote:
Hi Ola, all,
On 05/27/2015 10:08 AM, Ola Liljedahl wrote:
Here is my third attempt at a ODP API for MSGIO - asynchronous message
passing
based IPC for a shared nothing architecture.
I understand the value of this, but I feel
On Tue, Jun 02, 2015 at 03:08:55PM +0200, Nicolas Morey-Chaisemartin wrote:
Signed-off-by: Nicolas Morey-Chaisemartin nmo...@kalray.eu
---
v2:
* Change default value of rx_batch_len to the same as tx_batch_len
* Use odp_schedule_multi when Rx batching is enabled
https://bugs.linaro.org/show_bug.cgi?id=1608
Bug ID: 1608
Summary: api: add note about odp_pool_destroy
Product: OpenDataPlane
Version: unspecified
Hardware: Other
OS: Linux
Status: UNCONFIRMED
Severity:
https://bugs.linaro.org/show_bug.cgi?id=1603
Nicolas Morey-Chaisemartin nmo...@kalray.eu changed:
What|Removed |Added
CC||nmo...@kalray.eu
Signed-off-by: Nicolas Morey-Chaisemartin nmo...@kalray.eu
---
v2:
* Change default value of rx_batch_len to the same as tx_batch_len
* Use odp_schedule_multi when Rx batching is enabled
test/performance/odp_pktio_perf.c | 77 ++-
1 file changed, 59
On 2 June 2015 at 14:36, Benoît Ganne bga...@kalray.eu wrote:
Hi Ola, all,
On 05/27/2015 10:08 AM, Ola Liljedahl wrote:
Here is my third attempt at a ODP API for MSGIO - asynchronous message
passing
based IPC for a shared nothing architecture.
I understand the value of this, but I feel
On 2 June 2015 at 14:53, Jerin Jacob jerin.ja...@caviumnetworks.com wrote:
On Tue, Jun 02, 2015 at 02:36:47PM +0200, Benoît Ganne wrote:
Hi Ola, all,
On 05/27/2015 10:08 AM, Ola Liljedahl wrote:
Here is my third attempt at a ODP API for MSGIO - asynchronous message
passing
based IPC
UberConference Reminder___
lng-odp mailing list
lng-odp@lists.linaro.org
https://lists.linaro.org/mailman/listinfo/lng-odp
On 2 June 2015 at 16:46, Benoît Ganne bga...@kalray.eu wrote:
Thinking about using notification events instead of the messages
themselves...
If I use D-BUS or ZeroMQ, I get a socket descriptor. I don't want to
block on this descriptor, my application is likely blocking in
odp_schedule() or
Patch looks good. Mike I think you don't have objections to apply it, right?
Maxim.
On 05/28/15 09:24, Jerin Jacob wrote:
On Wed, May 27, 2015 at 07:38:26AM -0400, Mike Holmes wrote:
On 27 May 2015 at 00:17, Jacob, Jerin jerin.ja...@caviumnetworks.com
wrote:
Not completely understood your
Signed-off-by: Nicolas Morey-Chaisemartin nmo...@kalray.eu
---
v3:
* Fix help for rxbatch option
test/performance/odp_pktio_perf.c | 77 ++-
1 file changed, 59 insertions(+), 18 deletions(-)
diff --git a/test/performance/odp_pktio_perf.c
On 2 June 2015 at 11:22, Stuart Haslam stuart.has...@linaro.org wrote:
On Thu, May 28, 2015 at 12:58:08PM +0200, Christophe Milard wrote:
Renaming of things which may be, one day, exported in a lib.
This renaming is important, as it creates consistency between test
symbols, which is needed
Signed-off-by: Maxim Uvarov maxim.uva...@linaro.org
---
scripts/checkpatch.pl | 2 ++
1 file changed, 2 insertions(+)
diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl
index bc13831..ca60292 100755
--- a/scripts/checkpatch.pl
+++ b/scripts/checkpatch.pl
@@ -4255,6 +4255,8 @@ sub process
On Tue, Jun 02, 2015 at 04:57:45PM +0200, Nicolas Morey-Chaisemartin wrote:
Signed-off-by: Nicolas Morey-Chaisemartin nmo...@kalray.eu
Reviewed-by: Stuart Haslam stuart.has...@linaro.org
---
v3:
* Fix help for rxbatch option
test/performance/odp_pktio_perf.c | 77
Hi,
Many thanks Maciej for the insights! DPDK also does this TX completion
in the send call, but as Ola explained, this problem arise when you
can't expect that call to happen. E.g. because there is nothing new to
send, but you can't receive as well because all the buffers are waiting
to be
On 2 June 2015 at 11:31, Stuart Haslam stuart.has...@linaro.org wrote:
On Thu, May 28, 2015 at 12:58:10PM +0200, Christophe Milard wrote:
Module pktio now gets its own directory and create its own lib
(currentely only containing its executable)
Startup scripting stuff is just moved to the
On 1 June 2015 at 17:49, Maxim Uvarov maxim.uva...@linaro.org wrote:
Hello Christophe,
I just added few tests cases which we tested with Ciprian for current
version. I think you don't need to remove
from PATH original directories. Comments are:
1.
I don't like that name. _main is not
I dont have any objections.
On 2 June 2015 at 11:19, Maxim Uvarov maxim.uva...@linaro.org wrote:
Patch looks good. Mike I think you don't have objections to apply it,
right?
Maxim.
On 05/28/15 09:24, Jerin Jacob wrote:
On Wed, May 27, 2015 at 07:38:26AM -0400, Mike Holmes wrote:
On 27
Hi,
I have stated the reasoning in couple of weekly calls and this cover letter:
Implementations may be HW limited on how many worker threads can be supported
(efficiently). Number of control threads may be only SW limited. When user
identifies worker threads (from control threads), an
57 matches
Mail list logo