On 12/17/2015 08:56, HePeng wrote:
Hi, Maxim
I will present a quick patch that will not pass the kernel style
check as this one is just for evaluation.
ok, in that case you can add RFC to patch tag.
Maxim.
在 2015年12月16日,下午8:22,Maxim Uvarov 写道:
From meeting we
> -Original Message-
> From: EXT Zoltan Kiss [mailto:zoltan.k...@linaro.org]
> Sent: Wednesday, December 16, 2015 6:42 PM
> To: Savolainen, Petri (Nokia - FI/Espoo) ; lng-
> o...@lists.linaro.org; Elo, Matias (Nokia - FI/Espoo) ;
> Elo,
>
It’s better to allow packet output from any packet type pool, no matter when
the pool was created. Implementation controls packet pool memory mapping and
can e.g. arrange all packet pool memory to be contiguous (in global init time).
Each packet pool create could update IOMMU or if entries are
Petri,
Please also review API change.
On 11/17/2015 10:06, Balasubramanian Manoharan wrote:
Adds api to get packet error flags for L2, L3 and L4 errors.
Signed-off-by: Balasubramanian Manoharan
---
include/odp/api/packet_flags.h| 33
> -Original Message-
> From: EXT Zoltan Kiss [mailto:zoltan.k...@linaro.org]
> Sent: Wednesday, December 16, 2015 4:58 PM
> To: Savolainen, Petri (Nokia - FI/Espoo); lng-odp@lists.linaro.org
> Subject: Re: [lng-odp] [API-NEXT PATCH v5 2/7] api: pktio: added multiple
> pktio input queues
Merged,
Maxim.
On 12/08/2015 17:21, Ola Liljedahl wrote:
On 3 December 2015 at 12:37, Ivan Khoronzhuk
> wrote:
Timer API requires event types instead of buffer types.
Nit: this changes the linux-generic implementation, not
> -Original Message-
> From: EXT Zoltan Kiss [mailto:zoltan.k...@linaro.org]
> Sent: Wednesday, December 16, 2015 5:48 PM
> To: Savolainen, Petri (Nokia - FI/Espoo); lng-odp@lists.linaro.org
> Subject: Re: [lng-odp] [API-NEXT PATCH v5 2/7] api: pktio: added multiple
> pktio input queues
From: Alexandru Badicioiu
Measures the time required to launch and get the result
of ODP crypto operations in async and sync API mode with
configurable payload size, crypto algorithms and iteration
number.
Both asynchronous scheduled and polled are supported.
In scheduled mode a separate worker
Hi, Bala
Just pay attention, see below.
On 16.12.15 11:28, Balasubramanian Manoharan wrote:
Replaces odp_cos_create() function with odp_cls_cos_create() function
Signed-off-by: Balasubramanian Manoharan
---
test/validation/classification/classification.h| 1
Now redundant with atomic, barrier and lock tests.
Signed-off-by: Christophe Milard
---
configure.ac |1 -
platform/linux-generic/test/Makefile.am|1 -
test/validation/Makefile.am|
I will correct this as a separate patch.
Regards,
Bala
On 17 December 2015 at 15:48, Ivan Khoronzhuk
wrote:
> .
>
>>> +
>>> +pkt = receive_packet(, ODP_TIME_SEC_IN_NS);
>>
>> here has to be time of scheduler, ns should be converted to with
>>
https://bugs.linaro.org/show_bug.cgi?id=1921
--- Comment #2 from Bill Fischofer ---
Merged. Commit ID 7afa127e124b298dad557f192e02468cd9bcc5cc
--
You are receiving this mail because:
You are on the CC list for the bug.___
[API-NEXT/PATCHv2 1/6] api: classification: add class of serivce create api
^^^
typo can be fixed during merge: serivce -> service
Reviewed-by: Petri Savolainen
> -Original Message-
> From:
On 17.12.15 15:52, Savolainen, Petri (Nokia - FI/Espoo) wrote:
On Tuesday call, there were suggestions for another type of wait calls
/*
* Wait the specified number of nanoseconds
*/
void odp_time_wait_ns(uint64_t nsec);
/*
* Wait until the specified (wall clock) time has been
No functionnal changes: just code reordering to match the ODP modules.
Signed-off-by: Christophe Milard
---
configure.ac|3 +
platform/linux-generic/test/Makefile.am |3 +
test/validation/Makefile.am |5 +-
Hi
Before, we could find the implementation of odp-netmap on the opendataplane.org
but now I noticed there is only "linux-generic (netmap)", so I guess odp-netmap
has merged with linux-generic ? The git of odp-netmap seems not be uptated for
a while, the same for odp-keystone2. Does somebody
https://bugs.linaro.org/show_bug.cgi?id=1830
--- Comment #3 from Bala Manoharan ---
V2 needs to be merged after the following patch for class of service create
gets merged to API-NEXT https://patches.linaro.org/58539/
--
You are receiving this mail because:
You are
The question is related to the DMA mapping:
The driver needs to perform the DMA mapping, and we don't want to set up a
new DMA mapping at TX time for each individual packet: We want to set-up
the DMA at pktio_open() time, I guess, so that the NIC DMA can fetch any TX
packet to be transmitted in
On Tuesday call, there were suggestions for another type of wait calls
/*
* Wait the specified number of nanoseconds
*/
void odp_time_wait_ns(uint64_t nsec);
/*
* Wait until the specified (wall clock) time has been reached
*/
void odp_time_wait_until(odp_time_t time);
These two could
Merged,
Maxim.
On 12/17/2015 02:49, Bill Fischofer wrote:
For this series:
Reviewed-and-Tested-by: Bill Fischofer >
On Wed, Dec 16, 2015 at 7:45 AM, Petri Savolainen
isn't it defeating the purpose of the iommu to DMA map all possible mem
that could be used in a pool?... And as long as the memory is not mapped in
the user virtual space, vfio won't let me do that...
On 17 December 2015 at 14:24, Alexandru Badicioiu <
alexandru.badici...@linaro.org> wrote:
>
https://bugs.linaro.org/show_bug.cgi?id=1937
Bala Manoharan changed:
What|Removed |Added
Resolution|--- |FIXED
Why the driver must map pools one by one? Is this a (legacy) kernel driver?
The driver can map one or more DMA-ble memory regions at initialization
time and this memory region should be used for all PACKET type pools.
Alex
On 17 December 2015 at 15:15, Christophe Milard <
Yes, linux-generic has now a days Netmap support. You can activate netmap pktio
with ./configure script
./configure --with-netmap-path=
See DEPENDENCIES file for details.
-Petri
From: lng-odp [mailto:lng-odp-boun...@lists.linaro.org] On Behalf Of EXT Kury
Nicolas
Sent: Thursday, December
This series adds odp_time_local_res() and odp_time_wait_ns(),
odp_time_wait_until() calls to time API and it's validation
tests and usage.
Since v2:
- replaced odp_time_wait() on odp_time_wait_ns() and odp_time_wait_until()
- corrected validation tests to validate new wait APIs
Since v1:
- used
This patch adds validation test for odp_time_wait_ns/until().
Also it visually demonstrates 5 second count.
Signed-off-by: Ivan Khoronzhuk
---
test/validation/time/time.c | 53 +
test/validation/time/time.h | 2 ++
2 files
https://bugs.linaro.org/show_bug.cgi?id=1946
Mike Holmes changed:
What|Removed |Added
CC||mike.hol...@linaro.org
This function is used to get resolution, test uses it to
check other APIs.
Signed-off-by: Ivan Khoronzhuk
---
test/validation/time/time.c | 46 ++---
test/validation/time/time.h | 1 +
2 files changed, 32 insertions(+), 15
Signed-off-by: Ivan Khoronzhuk
---
test/validation/time/time.c | 4
1 file changed, 4 insertions(+)
diff --git a/test/validation/time/time.c b/test/validation/time/time.c
index 4040b9f..cef 100644
--- a/test/validation/time/time.c
+++
On 17/12/15 00:24, Bill Fischofer wrote:
On Tue, Dec 15, 2015 at 6:58 PM, Zoltan Kiss > wrote:
Hi,
I have a question: is it allowed to call these functions from worker
threads? The current linux-generic implementation
Ping
-Original Message-
From: ion.grig...@freescale.com [mailto:ion.grig...@freescale.com]
Sent: Monday, December 14, 2015 12:52 PM
To: lng-odp@lists.linaro.org
Cc: Grigore Ion-B17953
Subject: [PATCHv7] helper : Fix UDP checksum computation
From: Grigore Ion
https://bugs.linaro.org/show_bug.cgi?id=1921
Mike Holmes changed:
What|Removed |Added
Status|IN_PROGRESS |RESOLVED
Test checks scheduler time correctness using time API.
Signed-off-by: Ivan Khoronzhuk
---
Patch is based on api-next, as it uses new time API.
test/validation/scheduler/scheduler.c | 47 +--
1 file changed, 45 insertions(+), 2
Now odp_time_wait_ns() can be used instead of own implementation.
Signed-off-by: Ivan Khoronzhuk
---
test/performance/odp_pktio_perf.c | 16 +---
1 file changed, 1 insertion(+), 15 deletions(-)
diff --git a/test/performance/odp_pktio_perf.c
Signed-off-by: Ivan Khoronzhuk
---
test/validation/time/time.c | 4
1 file changed, 4 insertions(+)
diff --git a/test/validation/time/time.c b/test/validation/time/time.c
index 4040b9f..cef 100644
--- a/test/validation/time/time.c
+++
Signed-off-by: Ivan Khoronzhuk
---
test/validation/time/time.c | 10 ++
test/validation/time/time.h | 1 +
2 files changed, 11 insertions(+)
diff --git a/test/validation/time/time.c b/test/validation/time/time.c
index 50d9bdc..4040b9f 100644
---
This patch adds odp_time_local_res() and odp_time_wait_until(),
odp_time_wait_ns() APIs.
Signed-off-by: Ivan Khoronzhuk
---
include/odp/api/time.h| 21
platform/linux-generic/odp_time.c | 100 ++
2 files
Signed-off-by: Ivan Khoronzhuk
---
test/validation/time/time.c | 10 ++
test/validation/time/time.h | 1 +
2 files changed, 11 insertions(+)
diff --git a/test/validation/time/time.c b/test/validation/time/time.c
index 50d9bdc..4040b9f 100644
---
Now odp_time_wait_ns() can be used instead of own implementation.
Signed-off-by: Ivan Khoronzhuk
---
test/performance/odp_pktio_perf.c | 16 +---
1 file changed, 1 insertion(+), 15 deletions(-)
diff --git a/test/performance/odp_pktio_perf.c
This patch adds validation test for odp_time_wait_ns/until().
Also it visually demonstrates 5 second count.
Signed-off-by: Ivan Khoronzhuk
---
test/validation/time/time.c | 52 +
test/validation/time/time.h | 2 ++
2 files
v4 is present
On 17.12.15 17:29, Ivan Khoronzhuk wrote:
This series adds odp_time_local_res() and odp_time_wait_ns(),
odp_time_wait_until() calls to time API and it's validation
tests and usage.
Since v2:
- replaced odp_time_wait() on odp_time_wait_ns() and odp_time_wait_until()
- corrected
Test checks scheduler time correctness using time API.
Signed-off-by: Ivan Khoronzhuk
---
Based on api-next as it uses new time API.
Since v1:
- improved a little
test/validation/scheduler/scheduler.c | 45 +--
1 file changed, 43
v2 is present
On 17.12.15 18:20, Ivan Khoronzhuk wrote:
Test checks scheduler time correctness using time API.
Signed-off-by: Ivan Khoronzhuk
---
Patch is based on api-next, as it uses new time API.
test/validation/scheduler/scheduler.c | 47
This function is used to get resolution, test uses it to
check other APIs.
Signed-off-by: Ivan Khoronzhuk
---
test/validation/time/time.c | 46 ++---
test/validation/time/time.h | 1 +
2 files changed, 32 insertions(+), 15
This patch adds odp_time_local_res() and odp_time_wait_until(),
odp_time_wait_ns() APIs.
Signed-off-by: Ivan Khoronzhuk
---
include/odp/api/time.h| 21
platform/linux-generic/odp_time.c | 100 ++
2 files
This series adds odp_time_local_res() and odp_time_wait_ns(),
odp_time_wait_until() calls to time API and it's validation
tests and usage.
Since v3:
- improved validation test a little
Since v2:
- replaced odp_time_wait() on odp_time_wait_ns() and odp_time_wait_until()
- corrected validation
On 17 December 2015 at 09:53, Kury Nicolas
wrote:
> Hi
>
>
> Before, we could find the implementation of odp-netmap on the
> opendataplane.org but now I noticed there is only "linux-generic
> (netmap)", so I guess odp-netmap has merged with linux-generic ? The git
On Fri, Dec 11, 2015 at 05:10:51PM +, Stuart Haslam wrote:
> Modify the tests that currently transmit packets in batches to also
> receive packets in batches. This adds coverage of odp_queue_deq_multi()
> and odp_schedule_multi() specifically against a packet input queue,
> as this doesn't get
On Wed, Dec 16, 2015 at 03:45:34PM +0200, Petri Savolainen wrote:
> From: Matias Elo
>
> Added a separate function for determining if the netmap link
> is up.
>
> Signed-off-by: Matias Elo
> ---
> platform/linux-generic/pktio/netmap.c | 48
>
On 17/12/15 08:09, Elo, Matias (Nokia - FI/Espoo) wrote:
-Original Message-
From: EXT Zoltan Kiss [mailto:zoltan.k...@linaro.org]
Sent: Wednesday, December 16, 2015 6:42 PM
To: Savolainen, Petri (Nokia - FI/Espoo) ; lng-
o...@lists.linaro.org; Elo, Matias
On 17/12/15 08:54, Savolainen, Petri (Nokia - FI/Espoo) wrote:
-Original Message-
From: EXT Zoltan Kiss [mailto:zoltan.k...@linaro.org]
Sent: Wednesday, December 16, 2015 5:48 PM
To: Savolainen, Petri (Nokia - FI/Espoo); lng-odp@lists.linaro.org
Subject: Re: [lng-odp] [API-NEXT
On 17/12/15 08:21, Savolainen, Petri (Nokia - FI/Espoo) wrote:
-Original Message-
From: EXT Zoltan Kiss [mailto:zoltan.k...@linaro.org]
Sent: Wednesday, December 16, 2015 4:58 PM
To: Savolainen, Petri (Nokia - FI/Espoo); lng-odp@lists.linaro.org
Subject: Re: [lng-odp] [API-NEXT
https://bugs.linaro.org/show_bug.cgi?id=1946
--- Comment #1 from Bill Fischofer ---
Maxim: Can you be a bit more specific about what the problem is and how to
reproduce it? Thanks.
--
You are receiving this mail because:
You are on the CC list for the
Ok, then the patch should be reworked to do this. No point in dragging out
the process through a series of disruptive changes.
On Thu, Dec 17, 2015 at 5:56 AM, Savolainen, Petri (Nokia - FI/Espoo) <
petri.savolai...@nokia.com> wrote:
> Yes, that’s the end goal of this process. All
[PATCH] -> [API-NEXT PATCH]
> -Original Message-
> From: lng-odp [mailto:lng-odp-boun...@lists.linaro.org] On Behalf Of EXT
> Balasubramanian Manoharan
> Sent: Tuesday, November 17, 2015 9:06 AM
> To: lng-odp@lists.linaro.org
> Subject: [lng-odp] [PATCH] api: packet: add detailed packet
On 17 December 2015 at 17:31, Bill Fischofer wrote:
> Ok, then the patch should be reworked to do this. No point in dragging out
> the process through a series of disruptive changes.
Yes. The ultimate goal is to add this prefix to all existing APIs.
This patch only
If we want to split this into multiple patches for review convenience
that's fine. I'd just like to be sure that we don't have a series of API
breaks for 1.6, 1.7, etc. Best to do that all at once on a release
boundary.
On Thu, Dec 17, 2015 at 6:25 AM, Bala Manoharan
BTW, subject to that caveat I did review and test this patch series so
Reviewed-and-Tested-by: Bill Fischofer
On Thu, Dec 17, 2015 at 6:26 AM, Bill Fischofer
wrote:
> If we want to split this into multiple patches for review convenience
>
Agreed. But this odp_cls_cos_create() API change is not just a rename
and it modifies the parameters to CoS create and hence it was better
to do this in a separate review process.
Regards,
Bala
On 17 December 2015 at 17:56, Bill Fischofer wrote:
> If we want to split
Thanks Bill.
@Petri: Can you please provide your reviewed-by to this API change.
Regards,
Bala
On 17 December 2015 at 17:58, Bill Fischofer wrote:
> BTW, subject to that caveat I did review and test this patch series so
>
> Reviewed-and-Tested-by: Bill Fischofer
From: Hongbo Zhang
All CPU frequency APIs will be moved as linux-generic internal interfaces,
this patch removes validation of odp_cpu_hz_max().
Signed-off-by: Hongbo Zhang
---
test/validation/system/system.c | 9 -
From: Hongbo Zhang
All CPU frequency APIs will be moved as linux-generic internal interfaces,
this patch removes validation of odp_cpu_hz_id().
Signed-off-by: Hongbo Zhang
---
test/validation/system/system.c | 17 -
From: Hongbo Zhang
All CPU frequency APIs will be moved as linux-generic internal interfaces,
so they should not be referenced here.
Signed-off-by: Hongbo Zhang
---
test/api_test/odp_common.c| 1 -
test/performance/odp_atomic.c | 1
From: Hongbo Zhang
All CPU frequency APIs are moved for linux-generic internal, this patch
moves the declaration of odp_cpu_hz_max_id().
Signed-off-by: Hongbo Zhang
---
include/odp/api/cpu.h| 11 ---
From: Hongbo Zhang
All CPU frequency APIs are moved for linux-generic internal, this patch
moves the declaration of odp_cpu_hz_max().
Signed-off-by: Hongbo Zhang
---
include/odp/api/cpu.h| 9 -
Instead of removing these calls, we decided in an ARCH call this week that
odp_cpu_hz*() calls can return 0 on failure (e.g. when could not read CPU hz on
a platform).
-Petri
> -Original Message-
> From: EXT hongbo.zh...@freescale.com [mailto:hongbo.zh...@freescale.com]
> Sent:
Hmm, there were changes of decision how to handle these APIs.
Here it is mentioned to delete.
https://collaborate.linaro.org/pages/viewpage.action?title=2015-12-16+ODP+ARCH=LNG
Do you mean now the final official decision is to keep them, right?
"return 0 on failure", this can be fixed soon, and
From: Hongbo Zhang
All CPU frequency APIs will be moved as linux-generic internal interfaces,
this patch removes validation of odp_cpu_hz().
Signed-off-by: Hongbo Zhang
---
test/validation/system/system.c | 9 -
From: Hongbo Zhang
All CPU frequency APIs will be moved as linux-generic internal interfaces,
so they should not be referenced here.
Signed-off-by: Hongbo Zhang
---
example/classifier/odp_classifier.c | 4 +---
From: Hongbo Zhang
All CPU frequency APIs are moved for linux-generic internal, this patch
moves the declaration of odp_cpu_hz_id().
Signed-off-by: Hongbo Zhang
---
include/odp/api/cpu.h| 11 ---
From: Hongbo Zhang
All CPU frequency APIs are moved for linux-generic internal, this patch
moves the declaration of odp_cpu_hz().
Signed-off-by: Hongbo Zhang
---
include/odp/api/cpu.h| 9 -
The rationale behind this is that classification API should follow the same,
prefixed naming convention than other APIs. CoS is a generally used term and
e.g. scheduler, TM, packet, etc APIs could define also APIs around Class of
Service. Here odp_cos_t is essentially a node in a
Yes, we'll keep them. I wasn't involved in the delete decision.
You can add "@retval 0 on failure" to the API spec.
/**
* Current CPU frequency in Hz
*
* Returns current frequency of this CPU
*
* @return CPU frequency in Hz
* @retval 0 on failure
*/
uint64_t odp_cpu_hz(void);
-Petri
>
I agree with that general convention, however a class of service is a first
class entity (like a pool). It's why we have odp_pool_create() and
parameterize that it's a buffer vs. packet pool instead of having separate
odp_buffer_pool_create() and odp_packet_pool_create() APIs. Using that
analogy
What e.g. scheduler or TM would do with a odp_cos_t which is associated with
queue, pool and drop_policy?
typedef struct odp_cls_cos_param {
odp_queue_t queue; /**< Queue associated with CoS */
odp_pool_t pool;/**< Pool associated with CoS */
odp_drop_e
Fair enough, but then it seems we then need to add the odp_cls prefix to a
lot more APIs and typedefs in the existing classification.h to be fully
consistent. Right now there's a confusing mix of those that do and do not
have this prefix.
On Thu, Dec 17, 2015 at 5:49 AM, Savolainen, Petri (Nokia
Yes, that’s the end goal of this process. All classification API calls, types,
defines, … are prefixed with odp_cls_.
-Petri
From: EXT Bill Fischofer [mailto:bill.fischo...@linaro.org]
Sent: Thursday, December 17, 2015 1:54 PM
To: Savolainen, Petri (Nokia - FI/Espoo)
Cc: Bala Manoharan; LNG
---
helper/Makefile.am | 2 +
helper/cuckoo_hash.c| 728 ++
helper/include/odp/helper/cuckoo_hash.h | 434 ++
helper/include/odp/helper/ring.h| 43 ++
helper/ring.c | 28 ++
78 matches
Mail list logo