FYI
-- Forwarded message --
From: Bala Manoharan
Date: 27 April 2016 at 18:30
Subject: Re: classification questions
To: Zoltan Kiss
Cc: "Savolainen, Petri (Nokia - FI/Espoo)"
Regards,
Bala
On 25 April 2016 at 21:11, Zoltan Kiss wrote:
> Hi,
>
> Thi
On 27 April 2016 at 18:31, Bala Manoharan wrote:
> FYI
>
>
> -- Forwarded message ------
> From: Bala Manoharan
> Date: 27 April 2016 at 18:30
> Subject: Re: classification questions
> To: Zoltan Kiss
> Cc: "Savolainen, Petri (Nokia - FI/Espoo)"
Reviewed-by: Balasubramanian Manoharan
On 28 April 2016 at 01:08, Barry Spinney wrote:
> This patch adds a new marking_colors_supported and marking_colors_needed
> arrays. Also splits the TOS marking function - which previously could
> specify ECN or DSCP marking into two separate calls. Also
Reviewed-by: Balasubramanian Manoharan
On 27 April 2016 at 05:47, Bill Fischofer wrote:
> Add touch tests for the following APIs:
> odp_packet_has_l2_error()
> odp_packet_has_l3_error()
> odp_packet_has_l4_error()
>
> Signed-off-by: Bill Fischofer
> ---
> test/validation/packet/packet.c | 15
On 27 April 2016 at 21:43, Zoltan Kiss wrote:
> Move release to _odp_packet_classifier(), because caller has no way to
> know if new_pkt were allocated. In that case there would be a double free
> on pkt, and new_pkt would be leaked.
>
I am little skeptical about this classifier module freeing u
On 28 April 2016 at 21:50, Zoltan Kiss wrote:
>
>
> On 28/04/16 10:29, Bala Manoharan wrote:
>
>>
>> On 27 April 2016 at 21:43, Zoltan Kiss > <mailto:zoltan.k...@linaro.org>> wrote:
>>
>> Move release to _odp_packet_classifier(), because
On 29 April 2016 at 18:28, Zoltan Kiss wrote:
>
>
> On 28/04/16 18:08, Bala Manoharan wrote:
>
>>
>> On 28 April 2016 at 21:50, Zoltan Kiss > <mailto:zoltan.k...@linaro.org>> wrote:
>>
>>
>>
>> On 28/04/16 10:29, Bala Manoharan wr
On 3 May 2016 at 19:26, Zoltan Kiss wrote:
>
>
> On 02/05/16 09:56, Bala Manoharan wrote:
>
>>
>> On 29 April 2016 at 18:28, Zoltan Kiss > <mailto:zoltan.k...@linaro.org>> wrote:
>>
>>
>>
>> On 28/04/16 18:08, Bala Manoharan wr
For the series: Reviewed-by: Balasubramanian Manoharan <
bala.manoha...@linaro.org>
On 5 May 2016 at 21:05, Zoltan Kiss wrote:
> Fixing up things in error handling, and harmonizing different codebases.
>
> Zoltan Kiss (3):
> linux-generic: classification: release the packet as soon as an error
Hi,
FYI. Went through these old discussions while updating Traffic manager
documentation and felt the technical discussion might benefit the group.
Regards,
Bala
-- Forwarded message --
From: Barry Spinney
Date: 15 April 2016 at 23:41
Subject: RE: Marking API
To: Bala Manoharan
Reviewed-by: Balasubramanian Manoharan
On 10 May 2016 at 07:25, Bill Fischofer wrote:
> The tick_buf_t struct may be larger than 16 bytes when a lock char is
> needed so correct the ODP_STATIC_ASSERT to reflect this. This addresses
> bug https://bugs.linaro.org/show_bug.cgi?id=2211 when compili
Do you have any specific changes or additions to crypto API? or any
specific use-case which might benefit from the additional API?
Regards,
Bala
On 10 May 2016 at 20:50, Nikhil Agarwal wrote:
> Hi All,
>
> Since we have multi flavor for packet Rx/Tx APIs on queues and scheduler,
> shall we have
n() API? Maybe you can send an RFC with the changes and
we can discuss on the RFC
Regards,
Bala
>
> Regards
> Nikhil
>
> On 10 May 2016 at 21:03, Bala Manoharan wrote:
>
>> Do you have any specific changes or additions to crypto API? or any
>> specific use-case w
On 11 May 2016 at 20:15, Nikhil Agarwal wrote:
> From: Nikhil Agarwal
>
> Signed-off-by: Nikhil Agarwal
> ---
> doc/users-guide/users-guide.adoc | 81
> +++-
> 1 file changed, 63 insertions(+), 18 deletions(-)
>
> diff --git a/doc/users-guide/users-guide.ado
Hi Nikhil,
I had some comments on V1 regarding session type.
If you agree can you please incorporate on next version.
Regards,
Bala
On 12 May 2016 at 22:22, Nikhil Agarwal wrote:
> Yes I do agree with mentioning only the API names here. I did that in v2
> of this patch.
>
> On 12 May 2016 at 2
There is a use-case in classification, where the packet can be classified
based on outer/inner vlan tag and for that specific use-case it is better
to store the vlan value in packet header field.
Regards,
Bala
On 12 May 2016 at 18:06, Matias Elo wrote:
> There is no way to read vlan tags in the
Mpps (odp_l2fwd, 64B packets, netmap/dpdk pktio).
>
Sure. If the performance boost is by this much makes sense to remove this
now.
Regards,
Bala
> -Matias
>
>
>
>
>
> *From:* EXT Bala Manoharan [mailto:bala.manoha...@linaro.org]
> *Sent:* Friday, May 13, 2016 9:39 AM
>
Hi,
I have incorporated all the comments in V2 except adding reference for
IEEE802.1Q
I am little skeptical about adding a wikipedia page as reference.
Regards,
Bala
On 15 May 2016 at 03:09, Bill Fischofer wrote:
>
>
> On Fri, May 13, 2016 at 12:23 AM, Balasubramanian Manoharan <
> bala.manoha
On 17 May 2016 at 01:56, Bill Fischofer wrote:
> On Mon, May 16, 2016 at 6:00 PM, Nikhil Agarwal
> wrote:
>
>> Signed-off-by: Nikhil Agarwal
>>
>
> Reviewed-and-tested-by: Bill Fischofer
>
>
Reviewed-by: Balasubramanian Manoharan
>> ---
>> doc/users-guide/users-guide.adoc | 87
>> ++
FYI
@Bill: Looks like you did a reply instead of reply-all :)
Regards,
Bala
-- Forwarded message --
From: Bill Fischofer
Date: 16 May 2016 at 20:20
Subject: Re: [PATCHv2] doc: users-guide: add packet marking documentation
To: Balasubramanian Manoharan
On Mon, May 16, 2016 a
On 20 May 2016 at 05:05, Bill Fischofer wrote:
>
>
> On Thu, May 19, 2016 at 4:23 AM, Balasubramanian Manoharan
> wrote:
>>
>> Adds documentation for packet drop policy and Error Class of service
>>
>> Signed-off-by: Balasubramanian Manoharan
>> ---
>> doc/users-guide/users-guide-cls.adoc | 29
On 20 May 2016 at 20:11, Bill Fischofer wrote:
>
>
> On Fri, May 20, 2016 at 5:48 AM, Bala Manoharan
> wrote:
>>
>> On 20 May 2016 at 05:05, Bill Fischofer wrote:
>> >
>> >
>> > On Thu, May 19, 2016 at 4:23 AM, Balasubramanian Manoharan
>
Hi,
Comments inline...
On 26 May 2016 at 20:05, Dragos Adrian Badea wrote:
> Hi,
>
> I'm trying to understand the relation between a tm system and a pktio.
>
> I was wondering if you could you please briefly explain few aspects:
>
> -Is there a 1:1 mapping between a tm system and a pktio
ns the pktio in question has defined
> a single pktout at creation time ?
Yes. To make is more clear there cannot be more than one TM system
attached to a single pktio interface.
Regards,
Bala
>
> Regards,
> Dragos
>
>
> -----Original Message-
> From: Bala Manoharan [m
On 2 June 2016 at 20:37, Nikhil Agarwal wrote:
> Signed-off-by: Nikhil Agarwal
> ---
> include/odp/api/spec/crypto.h | 25 +++--
> 1 file changed, 23 insertions(+), 2 deletions(-)
>
> diff --git a/include/odp/api/spec/crypto.h b/include/odp/api/spec/crypto.h
> index d8123e9..
On 3 June 2016 at 15:29, Nikhil Agarwal wrote:
>
>
> -Original Message-
> From: lng-odp [mailto:lng-odp-boun...@lists.linaro.org] On Behalf Of Bala
> Manoharan
> Sent: Friday, June 03, 2016 2:32 PM
> To: Nikhil Agarwal
> Cc: LNG ODP Mailman List
> Subject:
Hi,
Just realised that I haven't sent these comment. Sorry for the late feedback.
Comments inline...
Regards,
Bala
On 2 June 2016 at 13:15, Matias Elo wrote:
> Don't allocate new packets inside of the internal
> classifier helpers _odp_packet_cls_enq() and
> _odp_packet_classifier(). Instead,
Regards,
Bala
On 8 June 2016 at 12:24, Elo, Matias (Nokia - FI/Espoo)
wrote:
>
>
>> On 2 June 2016 at 13:15, Matias Elo wrote:
>> > Don't allocate new packets inside of the internal
>> > classifier helpers _odp_packet_cls_enq() and
>> > _odp_packet_classifier(). Instead, save destination queue
Regards,
Bala
On 8 June 2016 at 22:38, Nikhil Agarwal wrote:
> Signed-off-by: Nikhil Agarwal
> ---
> include/odp/api/spec/crypto.h | 29 +++--
> 1 file changed, 27 insertions(+), 2 deletions(-)
>
> diff --git a/include/odp/api/spec/crypto.h b/include/odp/api/spec/crypto
On 9 June 2016 at 14:56, Elo, Matias (Nokia - FI/Espoo)
wrote:
>> >> On 2 June 2016 at 13:15, Matias Elo wrote:
>> >> > Don't allocate new packets inside of the internal
>> >> > classifier helpers _odp_packet_cls_enq() and
>> >> > _odp_packet_classifier(). Instead, save destination queue to
>> >>
On 10 June 2016 at 15:00, Nikhil Agarwal wrote:
>
>
> On 9 June 2016 at 16:50, Bala Manoharan wrote:
>>
>> Regards,
>> Bala
>>
>>
>> On 8 June 2016 at 22:38, Nikhil Agarwal wrote:
>> > Signed-off-by: Nikhil Agar
Regards,
Bala
On 10 June 2016 at 16:37, Nikhil Agarwal wrote:
>
>
> -Original Message-
> From: lng-odp [mailto:lng-odp-boun...@lists.linaro.org] On Behalf Of Bala
> Manoharan
> Sent: Friday, June 10, 2016 4:30 PM
> To: Nikhil Agarwal
> Cc: LNG ODP Mailman List
Reviewed-by: Balasubramanian Manoharan
Regards,
Bala
On 13 June 2016 at 22:16, Zoltan Kiss wrote:
> And repurpose packet_set_len() for this, as it is no longer used.
> This was introduced by the following commit, and break compatibility with
> ODP-DPDK:
>
> bd18047a "linux-gen: pktio: don't all
Hi,
In the existing TM api the odp_pktout_queue_t is given as input
parameter by the application to configure a TM system, the following
are the drawback of this approach,
1). A TM system is configured per pktio interface and not pktout queue
within an interface.
2). With the existing approach th
Regards,
Bala
On 23 June 2016 at 16:13, Savolainen, Petri (Nokia - FI/Espoo)
wrote:
>
>
>> -Original Message-
>> From: Balasubramanian Manoharan [mailto:bala.manoha...@linaro.org]
>> Sent: Monday, June 20, 2016 2:02 PM
>> To: lng-odp@lists.linaro.org
>> Cc: spin...@mellanox.com; Savolain
Hi,
This series is split into three patch set for ease of review and needs
to be merged into single patch set before upstream.
Regards,
Bala
On 28 June 2016 at 17:35, Balasubramanian Manoharan
wrote:
> Replaces pktio interface as input to TM system instead of
> odp_pktout_queue_t.This creates
Reviewed-by: Balasubramanian Manoharan
Regards,
Bala
On 17 May 2016 at 21:42, Rizwan Ansari wrote:
> Removes odp_sync_stores() call for OCTEON, Octeon
> build will fail, As this api has been already
> replaced by odp_mb_full().
>
> Signed-off-by: Rizwan Ansari
> Reviewed-by: Petri Savolainen
This approach of using a pcap file might be useful to recreate an
error scenario but IMO it will be difficult to maintain different pcap
file for test cases.
Is this method proposed for validation testing or for flow testing?
Regards,
Bala
On 14 July 2016 at 18:42, Bill Fischofer wrote:
> On Th
Regards,
Bala
On 26 July 2016 at 09:18, Brian Brooks wrote:
> On 07/25 19:49:36, Bill Fischofer wrote:
>> Add ODP_THREAD_SERVICE as a third type of ODP thread along with related
>> cpumask APIs that allow cores to be designated for running service threads.
>> Service threads are intended to be u
On 26 July 2016 at 22:26, Brian Brooks wrote:
> On 07/26 10:44:17, Maxim Uvarov wrote:
>> On 07/26/16 05:56, Brian Brooks wrote:
>> > --- a/platform/linux-generic/arch/mips64/cpu_arch.h
>> > +++ b/platform/linux-generic/arch/mips64/cpu_arch.h
>> > @@ -7,6 +7,8 @@
>> > #ifndef ODP_PLAT_CPU_ARCH_H
On 26 July 2016 at 23:09, Brian Brooks wrote:
> On 07/26 22:38:20, Bala Manoharan wrote:
>> On 26 July 2016 at 22:26, Brian Brooks wrote:
>> > On 07/26 10:44:17, Maxim Uvarov wrote:
>> >> On 07/26/16 05:56, Brian Brooks wrote:
>> >> > --- a/
one minor question: Is this API mainly for debugging or is there any
other use case? Apart from this
Reviewed-by: Balasubramanian Manoharan
Regards,
Bala
On 2 August 2016 at 14:20, wrote:
> From: Xuelin Shi
>
> generate mac string representation (XX:XX:XX:XX:XX:XX) of odph_ethaddr_t
>
> Sig
Hi,
This thread is started to clarify few discrepancy in classification
definition regarding the configuration of defaultCoS. Pls note that it
has been established in the classification module that is it the
responsibility of the application to make sure there is no discrepancy
while creating the
; — Andy Bechtolsheim
>
>
> > -Original Message-
> > From: lng-odp-boun...@lists.linaro.org [mailto:lng-odp-
> > boun...@lists.linaro.org] On Behalf Of Ciprian Barbu
> > Sent: Tuesday, October 14, 2014 3:32 PM
> > To: Bala Manoharan
> > Cc
Hi,
odp_packet_offset_map() API should be used generally if the application
wants to read a specific byte in a packet or
maybe to read any byte in a custom protocol.
odp_packet_l2_map() and odp_packet_l3_map() will be used to get the start
address of l2 and l3 layer.
Regards,
Bala
On 29 October
h
>>a version of the prototypes typedef'd appropriate to the platform
>>
>> Is there a clean way to achieve this and still permit efficient inlining?
>>
>>
>>
>> On Fri, Oct 31, 2014 at 2:08 AM, Bala Manoharan <
>> bala.manoha...@lina
Hi,
This patch will have to be modified as we have decided to move the
segmentation support in packet header file.
I will provide an updated patch for the same.
Regards,
Bala
On 31 October 2014 20:24, Bill Fischofer wrote:
> This patch does not apply to the current repository. Is there a revi
We should be able to modify the suite once we define the additional API in
crypto to read the supported crypto operations in the underlying platform.
The test suite could be written in such a way to run only the sync or
async test cases for the specified platform based on the return from the
query
Yes. I only need get_pktio_entry() function to be available in header file
in-order to be able to call the same from classifier.
Will incorporate the change in next version.
Regards,
Bala
On 18 November 2014 19:23, Stuart Haslam wrote:
> On Mon, Nov 17, 2014 at 11:03:39AM +, Balasubramanian
Thanks Ciprian.
Regards,
Bala
On 19 November 2014 18:36, Ciprian Barbu wrote:
> Hi,
>
> I used this tool for sending packets in batch mode, used it during OVS
> work.
>
> Here is a link:
> http://packeth.sourceforge.net/packeth/Home.html
>
> I believe it's going to be really useful for your cla
I agree. The test should only consider whether all the buffer that have
been enqueued
have been despatched by the scheduler.
Fair scheduling cannot be tested using cunit, that should be taken during
performance evaluation if needed.
Regards,
Bala
On 21 November 2014 15:55, Taras Kondratiuk
wrote
Hi Maxim,
Can we have those as separate patches.
I would like to keep this patch as specific for linking pktio with
classification.
I have created this patch from the current odp tip.
Regards,
Bala
On 21 November 2014 17:37, Maxim Uvarov wrote:
> On 11/21/2014 11:53 AM, Balasubramanian Manohar
On 21 November 2014 17:41, Maxim Uvarov wrote:
> On 11/21/2014 11:53 AM, Balasubramanian Manoharan wrote:
>
>> +cos_t *pktio_select_cos(pktio_entry_t *entry, uint8_t *pkt_addr,
>> + odp_packet_hdr_t *pkt_hdr)
>> +{
>> + pmr_t *pmr;
>> + cos_t *cos = NULL;
>>
> No
November 2014 18:18, Maxim Uvarov wrote:
> On 11/21/2014 03:32 PM, Bala Manoharan wrote:
>
>> Hi Maxim,
>>
>> Can we have those as separate patches.
>> I would like to keep this patch as specific for linking pktio with
>> classification.
>> I have c
Few points,
* Inorder to check ordered state of buffers from second queue they should
be dequeued by a single thread
as scheduler will despatch the buffers from ORDERED queue in initial order
but more than one thread can get the buffer from the same queue at the same
time.
Begards,
Bala
On 21 No
gt; On Fri, Nov 21, 2014 at 11:30 AM, Taras Kondratiuk <
> taras.kondrat...@linaro.org> wrote:
>
>> On 11/21/2014 06:16 PM, Ciprian Barbu wrote:
>>
>>> On Fri, Nov 21, 2014 at 5:54 PM, Bala Manoharan
>>> wrote:
>>>
>>>>
>>>> Few points
Hi,
The concern I had with the above testing scenario is that IMO this scenario
does not match perfectly with an application usecase. The testing for
ORDERED queue in my opinion could be,
1. Dequeue work W1-W10 from ORDERED queue Q1
2. Enqueue work W1-W5 to ORDERED queue Q2
3. Enqueue work W6-W10
IMO what we require for testing is a loop-back/dummy interface and not a
default interface.
I wouldn't like to modify the behaviour of odp_pktio_open()
If the requirement is creation of an loop-back interface, we need to
discuss new API to create a loop-back/dummy interface.
Regards,
Bala
On 25 N
Yes. But for testing the basic pktio APIs which I believe is Maxim's
requirement it can be done using loopback interface (configuring a HW
interface as loop is an implementation detail) so that there is no need for
any HW configuration as such while running the validation suite.
Packet classifier
Hi Mike,
As you know these classification patches was submitted as being
interdependent on each other.
But sure I can redo this specific patch to be independently compilable unit.
Also it would be great if I could get review comments on classification
implementation patches so that I can fix any
gt; > On Mon, Nov 24, 2014 at 3:37 PM, Maxim Uvarov
> > wrote:
> >>
> >> On 11/25/2014 12:34 AM, Mike Holmes wrote:
> >>>
> >>> This no longer applies, can you rebase it ?
> >>
> >>
> >> git am -3
> >>
> >&g
On 25 November 2014 21:54, Maxim Uvarov wrote:
> Define special name for loop back interface. That interface
> can be used mostly for testing. Each implementation can interpret
> that loop0 to any other device.
>
> Signed-off-by: Maxim Uvarov
> ---
> platform/linux-generic/include/api/odp_packe
Ping
On 26 November 2014 at 11:07, Bala Manoharan
wrote:
> Hi,
>
> Pls let me know if there are any additional comments for this patch.
> If otherwise please provide reviewed-by
>
> Regards,
> Bala
>
> On 25 November 2014 03:24, Anders Roxell wrote:
>
>>
Hi,
I agree with Victor's concern, implementation needs a mechanism to know
what is the amount of valid memory available in the mac_addr pointer.
If I am not wrong the idea of defining the ODP_PKTIO_MAC_ADDR_MAX_LEN was
initially discussed but the same was dropped as there were concerns since
this
I agree the #defines are bit verbose in this patch but I tried to create
the hash defines to be compatible with the current syntax in ODP header
files.
Also we had a discussion today in ODP API scrum regarding including
/usr/include/linux/tcp.h header file directly in linux-generic
implementations
I have only a few minor comments for this patch.
One major query is that Currently this patch does not integrate
headroom/tailroom at segment level
is the feature available in 1.0?
if it is needed to be supported then we need some modifications in segment
handling.
Regards,
Bala
On Tue, Dec 02,
Hi Maxim,
This odph_tcp struct is used to read the values form an incoming packet as
part of parsing and classification and hence the packets is always in
Network Byte order. ODP application can use the inline functions
odp_be_to_cpu_xx() to convert to cpu native byte order.
Victor,
We will add t
On 4 December 2014 at 18:46, Maxim Uvarov wrote:
> On 12/04/2014 03:59 PM, Bala Manoharan wrote:
>
>> Hi Maxim,
>>
>> This odph_tcp struct is used to read the values form an incoming packet
>> as part of parsing and classification and hence the packets is always
Hi,
In case of synchronous crypto operation the completion event is need only
in case of an error hence the application can check the return code for
failure and get info from completion event buffer only during that
scenario. This will greatly optimize the performance in synchronous crypto
operat
On Wed, Dec 10, 2014 at 12:54:38PM +0200, Ciprian Barbu wrote:
> On Tue, Dec 9, 2014 at 7:53 PM, Bill Fischofer
> wrote:
> > If you've already received the packet there's not much point in classifying
> > it since the output of classification is the queue that the packet should be
> > sent to (and
On Wed, Dec 10, 2014 at 02:46:34PM +0200, Ciprian Barbu wrote:
> On Wed, Dec 10, 2014 at 2:03 PM, Bala Manoharan
> wrote:
> > On Wed, Dec 10, 2014 at 12:54:38PM +0200, Ciprian Barbu wrote:
> >> On Tue, Dec 9, 2014 at 7:53 PM, Bill Fischofer
> >> wrote:
> &g
Thanks Shmulik!
Reviewed-by: Bala Manoharan
On 11 January 2015 at 23:02, Bill Fischofer
wrote:
> Thanks!
>
> On Sun, Jan 11, 2015 at 4:15 AM, Shmulik Ladkani <
> shmulik.ladk...@gmail.com> wrote:
>
>> ODP_{BIG,LITTLE}_ENDIAN_BITFIELD were always defined, regardle
an Manoharan
> wrote:
>>
>> Ping.
>>
>> On Thursday 01 October 2015 08:13 PM, Mike Holmes wrote:
>>
>> Thanks - I added a bug for each case
>>
>> On 1 October 2015 at 03:05, Bala Manoharan
>> wrote:
>>>
>>> On 30 September
Hi Bill,
Agreed. Maybe we can merge this patch and then I will send a separate
patch to resolve the conflict so that both your patch and the conflict
patch could be merged together.
IMO, if this patch is acceptable we can merge this as it is now and
then I can send a conflict resolution patch to b
This patch needs to be merged after https://patches.linaro.org/54903/
Regards,
Bala
On 15 October 2015 at 11:45, Balasubramanian Manoharan
wrote:
> odp_pktio_param_init() API is used to initialize odp_pktio_param_t params
>
> Signed-off-by: Balasubramanian Manoharan
> ---
> test/validation/cla
Hi Ivan,
Thanks for pointing out the issues. Since this patch is merged I will
create a bug and add the missing points.
Pls provide your inputs on the comments.
On 15 October 2015 at 16:53, Ivan Khoronzhuk wrote:
> Hi, Bala
>
> Just compared this version with requirements for v2 and saw some mis
Hi Ivan,
Comments inline.
On 15 October 2015 at 19:58, Ivan Khoronzhuk wrote:
>
>
> On 15.10.15 16:39, Bala Manoharan wrote:
>>
>> Hi Ivan,
>>
>> Thanks for pointing out the issues. Since this patch is merged I will
>> create a bug and add the missing po
Hi Ivan,
Agreed. Splitting can be done.
Do you have any other comments apart from splitting?
Regards,
Bala
On 20 October 2015 at 15:52, Ivan Khoronzhuk wrote:
> Hi, Bala
>
> Propose to split this on 4 patches for clarity and history reason:
>
> 1 - destroy_inq() move in order to use as common f
I skipped this coz this as I just realised that since this was an
error packet and it might not be required for platforms to parse an
error packet and hence I removed the sequence number check for this
packet and just checked whether this packet was received on error CoS
queue.
Regards,
Bala
On 2
.10.15 15:14, Bala Manoharan wrote:
>>
>> I skipped this coz this as I just realised that since this was an
>> error packet and it might not be required for platforms to parse an
>> error packet and hence I removed the sequence number check for this
>> packet and
Reviewed-by: Balasubramanian Manoharan
On 26 October 2015 at 16:09, Nicolas Morey-Chaisemartin
wrote:
> Fixes build issue on arch where uint32_t != unsigned:
> example/classifier/odp_classifier.c:694:8: error:
> passing argument 3 of ‘parse_value’ from incompatible
> pointer type [-Wer
Reviewed-by: Balasubramanian Manoharan
On 29 October 2015 at 16:38, Maxim Uvarov wrote:
> Looks like merge to api-next, cherry-pick from api-next
> and merge master to api-next produced duplicate patch lines.
> Found that with inspecting branches with 'git diff master api-next'
>
> Signed-off-by
Reviewed-by: Balasubramanian Manoharan
On 22 October 2015 at 18:11, wrote:
> From: Grigore Ion
>
> odph_ipv4_csum_update should be used to update the checksum inside a pkt,
> as it is used in all the other examples and tests different from
> classification. Thus the prototype of the function s
This method of identifying ODP version using API is useful for finding
the implementation version while executing the binary. IMO we need to
additionally define a method to identify the version of the compiled
binary statically.
We can add a "#define Version " inside the
implementation file so tha
Hi Petri,
Why don't we add this hash parameter to CoS so that when the packet
arrives to a CoS it can be distributed based on the hash algorithm to
a number of queues. Also if a system does not support Classification
this hashing can be attached to the default CoS and in systems
supporting Classif
ion API.
>>
>> Alex
>>
>>
>> On 9 November 2015 at 11:11, Savolainen, Petri (Nokia - FI/Espoo)
>> wrote:
>>>
>>>
>>>
>>> > -Original Message-
>>> > From: EXT Bala Manoharan [mailto:bala.manoha...@linaro.org]
>>> > Sen
inside classification.
Regards,
Bala
On 9 November 2015 at 16:02, Alexandru Badicioiu
wrote:
>
>
> On 9 November 2015 at 12:26, Bala Manoharan
> wrote:
>>
>> The existing Classification infra structure supports this behaviour of
>> hashing after classification we have
Hi,
I went through the same issue as well..
In ODP we have defined that the dev-name is implementation specific
and hence I like the idea to add socket type to dev-name something
like "SOCKET_MMAP:eth0"/ "NETMAP: eth0" and this maps well with the
ODP definition for device name.
Regards,
Bala
On
Hi,
I went through the same issue as well..
In ODP we have defined that the dev-name is implementation specific
and hence I like the idea to add socket type to dev-name something
like "SOCKET_MMAP:eth0"/ "NETMAP: eth0" and this maps well with the
ODP definition for device name.
Regards,
Bala
On
Thanks for pointing out.
Will update in the next version.
Regards,
Bala
On 17 November 2015 at 18:46, Bill Fischofer wrote:
> This should be marked API-NEXT.
>
> On Tue, Nov 17, 2015 at 12:16 AM, Balasubramanian Manoharan
> wrote:
>>
>> odp_cls_cos_pool_set() api is used to configure packet poo
Hi,
Pls find comments inline..
On 19 November 2015 at 16:24, Stuart Haslam wrote:
> On Tue, Nov 17, 2015 at 11:46:51AM +0530, Balasubramanian Manoharan wrote:
>> Adds support for configuring packet pool to a class-of-service.
>> linux-generic packet parser is enhanced to parse a packet directly
On 19 November 2015 at 16:43, Stuart Haslam wrote:
> On Tue, Nov 17, 2015 at 11:46:52AM +0530, Balasubramanian Manoharan wrote:
>> Assigns a packet pool to CoS using odp_cls_cos_pool_set() api.
>>
>
> This tests that you can assign a pool to a CoS but since it uses the
> same default_pool it doesn
Hi Bill,
I just tried the build and it seems to work fine on the latest API-NEXT branch.
Did you mean it is not working after applying only part 2 of the patch
ignoring the rest? All the 5 patches are dependent on each other since
the changes are across two modules classification and pktio.
I ca
>
> On Mon, Nov 23, 2015 at 11:23 PM, Bala Manoharan
> wrote:
>>
>> Hi Bill,
>>
>> I just tried the build and it seems to work fine on the latest API-NEXT
>> branch.
>>
>> Did you mean it is not working after applying only part 2 of the patch
>
Hi Petri,
Comments inline...
On 27 November 2015 at 19:52, Petri Savolainen
wrote:
> User area is either persistent or non-persistent. Init is
> called only once for persistent user area. Init is additionally
> called for non-persistent area when needed (when ODP overwrites
> previously initiali
Yes. This issue was pointed out by Stuart as well.
I will add detailed test case to test pool allocation with Cos.
Regards,
Bala
On 30 November 2015 at 14:49, Ivan Khoronzhuk
wrote:
> Hi, Bala
>
> I didn't look deep enough in this validation test but seems it doesn't check
> pool assignments for
I have sent V4 correcting the issues pointed above.
Regards,
Bala
On 3 December 2015 at 14:49, Savolainen, Petri (Nokia - FI/Espoo)
wrote:
>
>
>> -Original Message-
>> From: lng-odp [mailto:lng-odp-boun...@lists.linaro.org] On Behalf Of EXT
>> Balasubramanian Manoharan
>> Sent: Thursday,
Looks like a previous commit has changed the Ip address format
As a work around pls provide IP address format in hexadecimal.
For eg "10.10.10.1" ==> "A.A.A.1".
Pls assign the bug to me.
Regards,
Bala
On 10 December 2015 at 18:28, Bill Fischofer wrote:
> Please open a bug for this. Thanks.
>
>
cimal.
>>For eg "10.10.10.1" ==> "A.A.A.1".
>
> Thank you it works! But I still have no match with ODP_PMR_UDP_DPORT=1234
> even if I use the hexadecimal notation 4D2 or 04D2.
>
>
>
> De : Bala Manoharan
Hi,
odp_classifier and odp_generator is failing in the master branch but
both are working in api-next branch.
Looks like the issue in classifier is a bug is in the queue module
which was fixed in api-next branch and this fix needs to be merged
into master branch.
Regards,
Bala
___
1 - 100 of 528 matches
Mail list logo
