Re: [lng-odp] [PATCH] linux-generic: netmap: improve single RX queue performance

2016-02-12 Thread Maxim Uvarov
Merged, Maxim. On 02/11/16 16:51, Tilli, Juha-Matti (Nokia - FI/Espoo) wrote: Hi, All it took was increasing the link timeout from 5 seconds to 10 seconds to get it to work. It is visibly faster to start and to stop now after this change, thus ensuring minimal user annoyance. I don't recall

Re: [lng-odp] [PATCH v3 API-NEXT 1/4] linux-generic: separate x86 ODP_CACHE_LINE_SIZE to its arch file

2016-02-12 Thread Maxim Uvarov
On 02/09/16 18:48, Savolainen, Petri (Nokia - FI/Espoo) wrote: -Original Message- From: lng-odp [mailto:lng-odp-boun...@lists.linaro.org] On Behalf Of EXT Anders Roxell Sent: Tuesday, February 09, 2016 12:42 PM To: hongbo.zh...@linaro.org Cc: lng-odp@lists.linaro.org Subject: Re:

Re: [lng-odp] [PATCH v3 API-NEXT 1/4] linux-generic: separate x86 ODP_CACHE_LINE_SIZE to its arch file

2016-02-12 Thread Christophe Milard
On 5 February 2016 at 09:56, wrote: > From: Hongbo Zhang > > Currently all ODP_CACHE_LINE_SIZE macros for different architectures are > held in one header file, they should be moved to their own arch file. > This patch moves ODP_CACHE_LINE_SIZE

[lng-odp] [API-NEXT PATCHv2] test: l2fwd: add option to give destination MAC

2016-02-12 Thread Juha-Matti Tilli
Added option (--dst_addr or -r) to l2fwd that allows the user to specify what destination MAC addresses to use per port. The MAC addresses are separated by commas and the number of MAC addresses, if given, must be equal to the number of interfaces. This makes it much easier to test l2fwd with

[lng-odp] [API-NEXT PATCH 2/5] linux-generic: drv: creation of the driver interface (drv)

2016-02-12 Thread Christophe Milard
Signed-off-by: Christophe Milard --- platform/linux-generic/include/odp/drv/README | 2 ++ 1 file changed, 2 insertions(+) create mode 100644 platform/linux-generic/include/odp/drv/README diff --git a/platform/linux-generic/include/odp/drv/README

[lng-odp] [API-NEXT PATCH 7/9] api: pktio: specify when queue config is called

2016-02-12 Thread Petri Savolainen
Specify that queue config can be called always. Those must be called when input/output is not disabled or done through TM. Signed-off-by: Petri Savolainen --- include/odp/api/spec/packet_io.h | 28

[lng-odp] [API-NEXT PATCH 4/5] drv: adding standard types

2016-02-12 Thread Christophe Milard
Adding standard types to the ODP driver interface (drv). Signed-off-by: Christophe Milard --- include/odp/drv/spec/std_types.h | 40 1 file changed, 40 insertions(+) create mode 100644 include/odp/drv/spec/std_types.h diff

Re: [lng-odp] [API-NEXT PATCH 1/5] drv: creation of the driver interface (drv)

2016-02-12 Thread Christophe Milard
On 12 February 2016 at 15:16, Tilli, Juha-Matti (Nokia - FI/Espoo) < juha-matti.ti...@nokia.com> wrote: > Hi, > > Is there a tabulator in the first line of the commit message (i.e. in the > email subject)? If so, I would remove it and replace with a space. Applies > to many other patches in this

[lng-odp] [API-NEXT PATCH] example: l2fwd_simple: add single-threaded l2fwd example

2016-02-12 Thread Juha-Matti Tilli
The ODP distribution is lacking a simple example how to receive and send packets. The l2fwd is overly complicated because it supports an arbitrary number of interfaces, bi-directional operation and an arbitrary number of threads. To remedy this situation, add l2fwd_simple which is a

[lng-odp] [API-NEXT PATCH 0/5] creating the driver interface

2016-02-12 Thread Christophe Milard
This patch series creates the driver interface and populate it with one single file (defining the standard types). It also updates the documentation (implementers guide). This patch series creates the driver interface and populate it with one single file (defining the standard types). It also

[lng-odp] [API-NEXT PATCH 1/5] drv: creation of the driver interface (drv)

2016-02-12 Thread Christophe Milard
The Driver Programming Interface (drv) is defined in include/odp_drv.h and include/odp/drv/spec directory. Signed-off-by: Christophe Milard --- include/odp/drv/spec/README | 4 include/odp_drv.h | 24 2 files changed, 28

[lng-odp] [API-NEXT PATCH 8/9] test: pktio_perf: port to use new pktio api

2016-02-12 Thread Petri Savolainen
Ported odp_pktio_perf test from old single queue pktio API to the new API. To minimize porting effort, configured to use only single queue (default config). Signed-off-by: Petri Savolainen --- test/performance/odp_pktio_perf.c | 85

[lng-odp] [API-NEXT PATCH 9/9] test: l2fwd: utilize pktio param defaults

2016-02-12 Thread Petri Savolainen
Simplify pktio param setup by utilizing default values. Signed-off-by: Petri Savolainen --- test/performance/odp_l2fwd.c | 10 ++ 1 file changed, 2 insertions(+), 8 deletions(-) diff --git a/test/performance/odp_l2fwd.c b/test/performance/odp_l2fwd.c index

Re: [lng-odp] [API-NEXT PATCH 1/5] drv: creation of the driver interface (drv)

2016-02-12 Thread Tilli, Juha-Matti (Nokia - FI/Espoo)
Hi, If you have verified that there is no tab, it might be a technical problem with my Outlook. The problem I’m seeing is this subject line: [lng-odp] [API-NEXT PATCH 2/5] linux-generic: drv: creation of the driver interface (drv) …where there appears to be multiple spaces in the

[lng-odp] [PATCH] linux-generic: netmap: bump up link wait timeout to 10 secs

2016-02-12 Thread Juha-Matti Tilli
In some environments, waiting only for 5 seconds for the link is not enough. The problem is that the link wait timeouts, and thus startup of the application fails. Ten seconds seems to be a safe compromise. It is high enough to work on my test system reliably, yet low enough that it shouldn't

[lng-odp] [API-NEXT PATCH 6/9] api: pktio: define default pktio modes

2016-02-12 Thread Petri Savolainen
Defined direct modes as defaults. Simplifies mode selection in application. Especially, output mode is usually direct. Signed-off-by: Petri Savolainen --- include/odp/api/spec/packet_io.h | 8 ++-- platform/linux-generic/odp_packet_io.c | 2 ++ 2 files

[lng-odp] [API-NEXT PATCH 3/5] doc: implementers-guide: adding drv interface

2016-02-12 Thread Christophe Milard
The driver interface structure (similar to the api) is described. Signed-off-by: Christophe Milard --- doc/implementers-guide/implementers-guide.adoc | 77 -- 1 file changed, 60 insertions(+), 17 deletions(-) diff --git

[lng-odp] [API-NEXT PATCH 5/5] linux-generic: drv: adding standard types

2016-02-12 Thread Christophe Milard
Signed-off-by: Christophe Milard --- include/odp_drv.h | 2 ++ platform/linux-generic/include/odp/drv/std_types.h | 42 ++ 2 files changed, 44 insertions(+) create mode 100644

[lng-odp] [API-NEXT PATCH 3/9] api: pktio: added pktio info structure

2016-02-12 Thread Petri Savolainen
Added info structure so that application can request pktio device name and various configuration parameters. Harmonized usage of pktio parameter 'name' instead of 'dev'. Signed-off-by: Petri Savolainen --- include/odp/api/spec/packet_io.h | 35

[lng-odp] [API-NEXT PATCH 4/9] linux-generic: pktio: use unique names for pktin event queues

2016-02-12 Thread Petri Savolainen
Use odp name space and unique names for queues created by implementation. Signed-off-by: Petri Savolainen --- platform/linux-generic/odp_packet_io.c | 8 ++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/platform/linux-generic/odp_packet_io.c

[lng-odp] [API-NEXT PATCH 2/9] api: sched: define default sched param values

2016-02-12 Thread Petri Savolainen
Defines which values queue param init set by default. Default queue type is PLAIN, but it minimizes application work if also sched params are init with well known defaults. PARALLEL is chosen as default since it provides the lowest level of features. Signed-off-by: Petri Savolainen

Re: [lng-odp] [API-NEXT PATCH 1/5] drv: creation of the driver interface (drv)

2016-02-12 Thread Mike Holmes
It looks ok to me in gmail if that helps On 12 February 2016 at 10:25, Christophe Milard < christophe.mil...@linaro.org> wrote: > I cannot see any strange character in this. If anyone see the same, or can > understand this problem, please let me know. > Christophe. > > On 12 February 2016 at

Re: [lng-odp] [API-NEXT PATCH 1/5] drv: creation of the driver interface (drv)

2016-02-12 Thread Christophe Milard
Juha wrote that he sees the problem on other patches as well. so probably an outlook issue. I have dumped the series from gmail and mutt and could not find anything wrong. Juha will come back to me if it turns out to be a patch problem, but as it looks now, it feels the problem lies on his side.

Re: [lng-odp] [API-NEXT PATCH 1/5] drv: creation of the driver interface (drv)

2016-02-12 Thread Christophe Milard
I cannot see any strange character in this. If anyone see the same, or can understand this problem, please let me know. Christophe. On 12 February 2016 at 15:47, Tilli, Juha-Matti (Nokia - FI/Espoo) < juha-matti.ti...@nokia.com> wrote: > Hi, > > > > If you have verified that there is no tab, it