e lock order.
>
> This addresses the aspect of Bug
> https://bugs.linaro.org/show_bug.cgi?id=1879
> relating to deadlock in unicore systems.
>
> Suggested-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
> Signed-off-by: Bill Fischofer <bill.fischo...@linaro.org>
>
On 11/03/2015 10:21 AM, Nicolas Morey-Chaisemartin wrote:
>
> On 11/02/2015 06:11 PM, Bill Fischofer wrote:
>> Enqueueing to ordered queues requires two locks (source and target
>> queues). Since any queue can be either source or target, queues do not
>> have a natural l
> have a natural locking hierarchy. Create one by using the address of
> the qentry as the lock order.
>
> This addresses the aspect of Bug
> https://bugs.linaro.org/show_bug.cgi?id=1879
> relating to deadlock in unicore systems.
>
> Suggested-by: Nicolas Morey-Chaisemartin
On 11/03/2015 12:52 PM, Nicolas Morey-Chaisemartin wrote:
>
> On 10/30/2015 10:24 AM, Petri Savolainen wrote:
>> Added command line option to disable packet error check. Error
>> check requires full packet parse. Max packet rate measurements
>> should be done
On 10/30/2015 10:24 AM, Petri Savolainen wrote:
> Added command line option to disable packet error check. Error
> check requires full packet parse. Max packet rate measurements
> should be done with minimal feature set.
>
> This change gives +10% packet rate increase in direct recv mode
> with
For the series:
Reviewed-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
On 10/30/2015 10:24 AM, Petri Savolainen wrote:
> Added command line option to disable packet error check. Error
> check requires full packet parse. Max packet rate measurements
> should be done with mini
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
example/ipsec/odp_ipsec.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/example/ipsec/odp_ipsec.c b/example/ipsec/odp_ipsec.c
index 2aaae1d..9950ccf 100644
--- a/example/ipsec/odp_ipsec.c
+++ b/example
Reviewed-by: Nicolas Morey-Chaisemartion
On 11/03/2015 04:00 PM, Stuart Haslam wrote:
> A call to odp_queue_param_init() is missing so uninitialized parameters
> are passed to odp_queue_create() leading to unpredictable results.
>
> Signed-off-by: Stuart Haslam
On 10/30/2015 10:27 AM, Maxim Uvarov wrote:
> According to API odp_pktio_close() can be called only for stopped
> pktio. So in odp_pktio_term_global try to stop it first then
> call close. Also check all returns codes.
> https://bugs.linaro.org/show_bug.cgi?id=1854
>
> Signed-off-by: Maxim
On 10/30/2015 09:33 PM, Bill Fischofer wrote:
> To avoid deadlock, especially on a single core, force an explicit
> yield while not holding either lock when attempting to acquire multiple
> locks for ordered queue processing. Also handle enqueues to self as in
> this case the origin and target
On 10/30/2015 12:46 PM, Bill Fischofer wrote:
> A few comments:
>
> On Fri, Oct 30, 2015 at 3:38 AM, Christophe Milard
> > wrote:
>
> Maybe time to move things around and embrace more than just the
> validation tests in the
On 10/30/2015 04:07 PM, Mike Holmes wrote:
> SNIP -
>
> I can speak from experience: there are a lot of reasons for tests to be
> "specific" to a platform.
>
> Here are some we had to deal with:
> - No cache coherency
> - No glibc (gettimeofday for example)
> - Incompatible
one pktio provided.
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
test/validation/pktio/pktio.c | 72 +++
1 file changed, 38 insertions(+), 34 deletions(-)
diff --git a/test/validation/pktio/pktio.c b/test/validation/pktio/pkti
Drop packets counts both packets dropped by calls to pktio_send()
and packet which were flagged with errors during parsing.
Add an extra counter to differ these two types of packets.
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
v2:
Rebased on latest master
counts both packets dropped by calls to pktio_send()
>> and packet which were flagged with errors during parsing.
>>
>> Add an extra counter to differ these two types of packets.
>>
>> Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
>> ---
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
Reviewed-by: Petri Savolainen <petri.savolai...@nokia.com>
---
test/validation/packet/packet.c | 81 +
test/validation/packet/packet.h | 1 +
2 files changed, 82 insertions(+)
diff
For the moment just a simple wrapper around buffer_alloc to be used by
odp_buffer_alloc_multi and odp_packet_alloc_multi
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
Reviewed-by: Petri Savolainen <petri.savolai...@nokia.com>
---
platform/linux-gen
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
Reviewed-by: Petri Savolainen <petri.savolai...@nokia.com>
---
include/odp/api/packet.h | 29 +
1 file changed, 29 insertions(+)
diff --git a/include/odp/api/packet.h b/include/odp/api/packet.h
i
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
Reviewed-by: Petri Savolainen <petri.savolai...@nokia.com>
---
test/validation/buffer/buffer.c | 106
test/validation/buffer/buffer.h | 2 +
2 files changed, 108 insertions(+)
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
Reviewed-by: Petri Savolainen <petri.savolai...@nokia.com>
---
platform/linux-generic/odp_pool.c | 15 +++
1 file changed, 15 insertions(+)
diff --git a/platform/linux-generic/odp_pool.c
b/platform/linux-generi
On 10/28/2015 09:30 AM, Savolainen, Petri (Nokia - FI/Espoo) wrote:
>
>> -Original Message-
>> From: EXT Nicolas Morey-Chaisemartin [mailto:nmo...@kalray.eu]
>> Sent: Wednesday, October 28, 2015 10:23 AM
>> To: Savolainen, Petri (Nokia - FI/Espoo);
On 10/28/2015 09:12 AM, Savolainen, Petri (Nokia - FI/Espoo) wrote:
>>> Nicolas, you could define a platform specific init struct that gives user
>> option to tell which u64 operations will be called. If user tells that
>> cas/min/max_u64 are not called, you can implement everything lock-free.
do not make asumptions on the amount of buffer returned
* Add more test coverage
* Free to multiple pools at once
* Alloc more than pool contains
* etc.
v2:
* Fix prototypes
* Add alloc_multi functions
* Add implementations
* Add tests
Nicolas Morey-Chaisemartin (7):
api: buffer:
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
Reviewed-by: Petri Savolainen <petri.savolai...@nokia.com>
---
include/odp/api/buffer.h | 26 ++
1 file changed, 26 insertions(+)
diff --git a/include/odp/api/buffer.h b/include/odp/api/buffer.h
i
all passed.
> Please take a look at it.
>
>
> Maxim.
>
> On 10/27/2015 20:52, Nicolas Morey-Chaisemartin wrote:
>> Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
>> ---
>> include/odp/api/buffer.h | 26 ++
>> 1 file ch
On 10/27/2015 04:16 PM, Ivan Khoronzhuk wrote:
>
>
> On 27.10.15 16:48, Petri Savolainen wrote:
>> Platforms may support some uint64 operations lock-free and
>> others not. For example, inc_64 can be natively supported but
>> cas_64 (or max_64/min_64) not. User may be able to switch to
>> 32 bit
On 10/27/2015 05:14 PM, Savolainen, Petri (Nokia - FI/Espoo) wrote:
>
>> -Original Message-
>> From: EXT Nicolas Morey-Chaisemartin [mailto:nmo...@kalray.eu]
>> Sent: Tuesday, October 27, 2015 5:35 PM
>> To: Ivan Khoronzhuk; Savolainen, Petri (N
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
platform/linux-generic/odp_pool.c | 15 +++
1 file changed, 15 insertions(+)
diff --git a/platform/linux-generic/odp_pool.c
b/platform/linux-generic/odp_pool.c
index 9117cc8..9859ff6 100644
--- a/platform/linux-g
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
include/odp/api/buffer.h | 26 ++
1 file changed, 26 insertions(+)
diff --git a/include/odp/api/buffer.h b/include/odp/api/buffer.h
index aea273f..6631f47 100644
--- a/include/odp/api/buffer.h
+++ b/i
nce
* Alloc more than pool contains
* etc.
v2:
* Fix prototypes
* Add alloc_multi functions
* Add implementations
* Add tests
Nicolas Morey-Chaisemartin (7):
api: buffer: add functions to alloc/free multiple buffers at once
api: packet: add functions to alloc/free multiple packets at o
For the moment just a simple wrapper around buffer_alloc to be used by
odp_buffer_alloc_multi and odp_packet_alloc_multi
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
platform/linux-generic/include/odp_buffer_internal.h | 2 ++
platform/linux-generic/odp_
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
test/validation/buffer/buffer.c | 106
test/validation/buffer/buffer.h | 2 +
2 files changed, 108 insertions(+)
diff --git a/test/validation/buffer/buffer.c b/test/validation/
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
test/validation/packet/packet.c | 81 +
test/validation/packet/packet.h | 1 +
2 files changed, 82 insertions(+)
diff --git a/test/validation/packet/packet.c b/test/validation/
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
platform/linux-generic/odp_packet.c | 28
1 file changed, 28 insertions(+)
diff --git a/platform/linux-generic/odp_packet.c
b/platform/linux-generic/odp_packet.c
index 1b49624..79bf8b3
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
include/odp/api/packet.h | 29 +
1 file changed, 29 insertions(+)
diff --git a/include/odp/api/packet.h b/include/odp/api/packet.h
index 5d46b7b..85cc9c9 100644
--- a/include/odp/api/packet.h
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 [-Werror] [policy_count].rule.val_sz);
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
e
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
platform/linux-generic/odp_queue.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/platform/linux-generic/odp_queue.c
b/platform/linux-generic/odp_queue.c
index de22bd6..a7022cf 100644
--- a/platform
Remove the call to odp_term_global when a thread returns.
This causes some serious side effects (and bugs) when thread are
created and joined "dynamically".
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
helper/linux.c | 2 --
1 file changed, 2 deletions(-)
dif
:04 PM, Nicolas Morey-Chaisemartin wrote:
> This allows pktio to support concurrent recv/send.
>
> For the current pktios:
> * loop uses a queue (serializing)
> * socket/sockmmsg are serialized by the socket itself
> * socket_mmap now handles concurrent accesses
> * lo
Platform specific m4 files cannot define AM_CONDITIONAL.
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
configure.ac | 1 +
platform/linux-generic/m4/odp_pcap.m4 | 1 -
2 files changed, 1 insertion(+), 1 deletion(-)
diff --git a/configur
Any more feedback on this ?
On 10/13/2015 10:47 AM, Nicolas Morey-Chaisemartin wrote:
> As a lot of ODP calls allows the user to receive/poll/send multiple packets
> at once, I think it makes sense to provide a call
> to free a bunch of packets/buffers at the same time.
>
obal when a thread returns.
>> This causes some serious side effects (and bugs) when thread are
>> created and joined "dynamically".
>>
>> Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
>> ---
>> helper/linux.c | 2 --
>> 1
ly and scales quite well with the number of cores.
>
>> -Original Message-
>> From: lng-odp [mailto:lng-odp-boun...@lists.linaro.org] On Behalf Of EXT
>> Nicolas Morey-Chaisemartin
>> Sent: Tuesday, October 20, 2015 6:05 PM
>> To: lng-odp@lists.linaro.org
>> Sub
On 10/21/2015 10:28 AM, Savolainen, Petri (Nokia - FI/Espoo) wrote:
>
>> -Original Message-
>> From: EXT Nicolas Morey-Chaisemartin [mailto:nmo...@kalray.eu]
>> Sent: Wednesday, October 21, 2015 11:03 AM
>> To: Savolainen, Petri (Nokia - FI/Espoo); lng-odp@list
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
platform/linux-generic/odp_packet_io.c | 10 +-
1 file changed, 9 insertions(+), 1 deletion(-)
diff --git a/platform/linux-generic/odp_packet_io.c
b/platform/linux-generic/odp_packet_io.c
index 8a09439..1245745
Ping
On 09/30/2015 11:10 AM, Maxim Uvarov wrote:
> looks reasonable. Does somebody have any objections?
>
> Maxim.
>
> On 09/22/15 10:59, Nicolas Morey-Chaisemartin wrote:
>> Drop packets counts both packets dropped by calls to pktio_send()
>> and packet which were
For legacy reason loop on all pktio ranged from 1 to
ODP_CONFIG_PKTIO_ENTRIES included.
Remove loop offset to simplify code
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
platform/linux-generic/odp_packet_io.c | 12 ++--
1 file changed, 6 insertions(+), 6 del
Only call odp_pktio_close on opened pktio and output an error
message if odp_pktio_close failed.
Fixes: https://bugs.linaro.org/show_bug.cgi?id=1854
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
platform/linux-generic/odp_packet_io.c | 25 ++---
in the ring should be flushed by the thread which reserved
the slot in the ring buffer in the first place.
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
platform/linux-generic/include/odp_packet_socket.h | 2 +-
platform/linux-generic/pktio/socket_mmap.c
one might
not be able to switch the promisc mode for example.
Ideally a rwlock which favors write would solve the issue.
I don't have a proper test setup but running pktio perf in polling mode with 4
threads on sockets showed a x2 speed up.
Nicolas Morey-Chaisemartin (2):
linux-generic
concurrently but only one thread each.
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
.../linux-generic/include/odp_packet_io_internal.h | 4 +-
platform/linux-generic/include/odp_packet_netmap.h | 3 +
platform/linux-generic/odp_packet_io.c
On 10/16/2015 02:39 PM, Mike Holmes wrote:
> Nicolas should you have added you sign-off since you added to our code base
> ?
I wasn't sure about the policy. Concerning checkpatch, I'd trust these people
signoff over mine ;)
Anyway maxim added his so it should be alright
; Maxim.
>
> On 10/14/2015 10:53, Nicolas Morey-Chaisemartin wrote:
>> There at still 3 loops with the same = 1, <=PKTIO_ENTRIES in the packet_io.c
>> file.
>> Can we keep this patch this way and then send a new patch that fixes all
>> these loops at once?
>>
Ping
On 09/14/2015 11:14 AM, Nicolas Morey-Chaisemartin wrote:
> Remove duplicated code and use preexisting metadata copy function
>
> Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
> ---
> platform/linux-generic/odp_packet.c | 34 ++-
Please forget the first two sentences of my previous mail.
I started writing an alternate patch and then realized it was fixed in master
already.
On 10/15/2015 05:20 PM, Nicolas Morey-Chaisemartin wrote:
> Maybe changing the directory before running git_hash.sh would fix your issue
> an
Git hash was modified to handled non git builds but when building
out of tree, it detects a .git folder but do not change to this
folder to run git command.
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
scripts/git_hash.sh | 4 ++--
1 file changed, 2 insertions(+), 2 del
Maybe changing the directory before running git_hash.sh would fix your issue
and keep the old make working ?
Something like
This was fixed some time ago:
commit 06537738ea438c3e339fc269dedb4ed6c5e48f07
Author: Anders Roxell
Date: Fri Aug 14 15:25:05 2015 +0200
nect and he was surprised.
> We tried several methods with modifing dirs.
> When I've used $() instead of `` it helped but broke smth else...
>
>
> On 15.10.15 18:20, Nicolas Morey-Chaisemartin wrote:
>> Maybe changing the directory before running git_hash.sh wou
There at still 3 loops with the same = 1, <=PKTIO_ENTRIES in the packet_io.c
file.
Can we keep this patch this way and then send a new patch that fixes all these
loops at once?
Nicolas
On 10/14/2015 09:52 AM, Maxim Uvarov wrote:
> On 10/14/2015 10:21, Nicolas Morey-Chaisemartin
On 10/14/2015 08:48 AM, Maxim Uvarov wrote:
> On 10/13/2015 20:05, Nicolas Morey-Chaisemartin wrote:
>> Right now, pktio_term calls term on all pktio type but
>> some pktio might still be be opened. Although the user should
>> probably close its pktio before termint
For the moment just a simple wrapper around buffer_alloc to be used by
odp_buffer_alloc_multi and odp_packet_alloc_multi
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
platform/linux-generic/include/odp_buffer_internal.h | 2 ++
platform/linux-generic/odp_
ons
* Add implementations
* Add tests
Nicolas Morey-Chaisemartin (7):
api: buffer: add functions to alloc/free multiple buffers at once
api: packet: add functions to alloc/free multiple packets at once
linux-generic: pool: add buffer_alloc_multi function
linux-generic: pool: add implementat
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
test/validation/buffer/buffer.c | 106
test/validation/buffer/buffer.h | 2 +
2 files changed, 108 insertions(+)
diff --git a/test/validation/buffer/buffer.c b/test/validation/
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
platform/linux-generic/odp_packet.c | 28
1 file changed, 28 insertions(+)
diff --git a/platform/linux-generic/odp_packet.c
b/platform/linux-generic/odp_packet.c
index 209a6e6..5a1e870
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
test/validation/packet/packet.c | 81 +
test/validation/packet/packet.h | 1 +
2 files changed, 82 insertions(+)
diff --git a/test/validation/packet/packet.c b/test/validation/
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
platform/linux-generic/odp_pool.c | 15 +++
1 file changed, 15 insertions(+)
diff --git a/platform/linux-generic/odp_pool.c
b/platform/linux-generic/odp_pool.c
index 9036147..d4f6f14 100644
--- a/platform/linux-g
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
include/odp/api/buffer.h | 26 ++
1 file changed, 26 insertions(+)
diff --git a/include/odp/api/buffer.h b/include/odp/api/buffer.h
index aea273f..6631f47 100644
--- a/include/odp/api/buffer.h
+++ b/i
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
include/odp/api/packet.h | 29 +
1 file changed, 29 insertions(+)
diff --git a/include/odp/api/packet.h b/include/odp/api/packet.h
index 5d46b7b..85cc9c9 100644
--- a/include/odp/api/packet.h
On 10/09/2015 01:59 PM, Maxim Uvarov wrote:
> Signed-off-by: Maxim Uvarov
> ---
> platform/linux-generic/Makefile.am | 2 +
> .../linux-generic/include/odp_buffer_internal.h| 3 +
> .../linux-generic/include/odp_packet_io_internal.h | 35 +
>
Right now, pktio_term calls term on all pktio type but
some pktio might still be be opened. Although the user should
probably close its pktio before termintating the application,
it is safer to iterate on all pktio and close them in the
pktio_term function.
Signed-off-by: Nicolas Morey
# ({...
$ctx !~
/^.\s*#\s*define\s+TRACE_(?:SYSTEM|INCLUDE_FILE|INCLUDE_PATH)\b/)
{
$ctx =~ s/\n*$//;
On 10/09/2015 09:44 AM, Nicolas Morey-Chaisemartin wrote:
> I guess it should. But I'm not sure where the upstream is.
> It may even be
single PMR and PMR sets
creation
Benoît Ganne (3):
api: classification: move odp_pmr_match_t definition
linux-generic: classification: implement ODP_PMR_CUSTOM_FRAME matching
example: classifier: implement ODP_PMR_CUSTOM_FRAME match
Nicolas Morey-Chaisemartin (2):
api: classification
Signed-off-by: Benoît Ganne
---
example/classifier/odp_classifier.c | 86 ++---
1 file changed, 70 insertions(+), 16 deletions(-)
diff --git a/example/classifier/odp_classifier.c
b/example/classifier/odp_classifier.c
index 689313d..fa10809
Signed-off-by: Benoît Ganne
---
.../include/odp_classification_datamodel.h | 2 ++
.../include/odp_classification_inlines.h| 21 +
platform/linux-generic/odp_classification.c | 7 +++
3 files changed, 30 insertions(+)
diff
I guess it should. But I'm not sure where the upstream is.
It may even be fixed there.
Nicolas
On 10/08/2015 06:20 PM, Mike Holmes wrote:
> Should this go upstream also ?
>
> On 8 October 2015 at 14:46, Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
> wrote:
>
>>
The application can now specify a packet offset
in PMR rules. This offset is absolute from the
frame start. It is used to extract the PMR value.
This is useful to support arbitrary backplane
protocols and extensions.
Signed-off-by: Benoît Ganne
---
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
example/classifier/odp_classifier.c| 12 +---
include/odp/api/classification.h | 31 +++
platform/linux-generic/odp_classification.c
This patch move the odp_pmr_match_t structure definition
to prepare the introduction of the new ODP_PMR_CUSTOM_FRAME
term.
Signed-off-by: Benoît Ganne
---
include/odp/api/classification.h | 30 +++---
1 file changed, 15 insertions(+), 15 deletions(-)
OK. Just submitted the cherry picked patch on the ML.
On 10/09/2015 02:13 PM, Maxim Uvarov wrote:
> On 10/09/15 12:21, Nicolas Morey-Chaisemartin wrote:
>> It is actually already fixed in the Linux kernel.
>> Shoud we import the latest checkpatch then?
>
> we should cherry-p
Using checkpatch.pl with Perl 5.22.0 generates the following warning:
Unescaped left brace in regex is deprecated, passed through in regex;
This patch fixes the warnings by escaping occurrences of the left brace
inside the regular expression.
Signed-off-by: Eddie Kovsky
On 09/01/2015 02:23 PM, Maxim Uvarov wrote:
> On 09/01/15 15:07, Nicolas Morey-Chaisemartin wrote:
>>
>> On 09/01/2015 12:39 PM, Maxim Uvarov wrote:
>>> On 09/01/15 11:45, Nicolas Morey-Chaisemartin wrote:
>>>> Hi,
>>>>
>>>>
On 10/09/2015 02:59 PM, Maxim Uvarov wrote:
> Add api for packet reference count support. Which is useful in case:
> - multicast support
> - TCP retransmission
> - traffic generator
>
> Introduced new call: newpkt = odp_packet_create_ref(pkt) which creates
> reference to original packet, but
the macro was only invoked conditionally.
Solve this by setting a default status do disabled/no at the top configure,
which can be overridden by the platform m4 files. And then set the
AM_CONDITIONAl in the top configure.ac for every platform to see.
Signed-off-by: Nicolas Morey-Chaisem
On 09/30/2015 12:47 PM, Savolainen, Petri (Nokia - FI/Espoo) wrote:
>
>> -Original Message-
>> From: lng-odp [mailto:lng-odp-boun...@lists.linaro.org] On Behalf Of EXT
>> Nicolas Morey-Chaisemartin
>> Sent: Tuesday, September 29, 2015 5:16 PM
>> To: ln
Ping
On 09/10/2015 05:22 PM, Nicolas Morey-Chaisemartin wrote:
> This allows better readability, and make supporting ordered queues
> easier for other platforms
>
> Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
> ---
> platform/linux-g
On 10/08/2015 12:43 PM, Savolainen, Petri (Nokia - FI/Espoo) wrote:
>
>> -Original Message-
>> From: EXT Nicolas Morey-Chaisemartin [mailto:nmo...@kalray.eu]
>> Sent: Thursday, October 08, 2015 12:43 PM
>> To: Savolainen, Petri (Nokia - FI/Espoo); lng-odp@list
* Test do not make asumptions on the amount of buffer returned
* Add more test coverage
* Free to multiple pools at once
* Alloc more than pool contains
* etc.
v2:
* Fix prototypes
* Add alloc_multi functions
* Add implementations
* Add tests
Nicolas Morey-Chaisemartin (7):
api
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
test/validation/buffer/buffer.c | 105
test/validation/buffer/buffer.h | 2 +
2 files changed, 107 insertions(+)
diff --git a/test/validation/buffer/buffer.c b/test/validation/
ed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
scripts/checkpatch.pl | 8
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl
index 4983a0c..893da37 100755
--- a/scripts/checkpatch.pl
+++ b/scripts/checkpatch.p
For the moment just a simple wrapper around buffer_alloc to be used by
odp_buffer_alloc_multi and odp_packet_alloc_multi
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
platform/linux-generic/include/odp_buffer_internal.h | 2 ++
platform/linux-generic/odp_
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
platform/linux-generic/odp_packet.c | 28
1 file changed, 28 insertions(+)
diff --git a/platform/linux-generic/odp_packet.c
b/platform/linux-generic/odp_packet.c
index 209a6e6..5a1e870
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
test/validation/packet/packet.c | 80 +
test/validation/packet/packet.h | 1 +
2 files changed, 81 insertions(+)
diff --git a/test/validation/packet/packet.c b/test/validation/
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
include/odp/api/buffer.h | 26 ++
1 file changed, 26 insertions(+)
diff --git a/include/odp/api/buffer.h b/include/odp/api/buffer.h
index aea273f..6631f47 100644
--- a/include/odp/api/buffer.h
+++ b/i
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
platform/linux-generic/odp_pool.c | 15 +++
1 file changed, 15 insertions(+)
diff --git a/platform/linux-generic/odp_pool.c
b/platform/linux-generic/odp_pool.c
index 9036147..d4f6f14 100644
--- a/platform/linux-g
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
include/odp/api/packet.h | 29 +
1 file changed, 29 insertions(+)
diff --git a/include/odp/api/packet.h b/include/odp/api/packet.h
index 5d46b7b..85cc9c9 100644
--- a/include/odp/api/packet.h
You could probably generate a convenience library with both merged as a
single.a/.la file ?
I feel it would also be better for an end user to have a single library if we
start having cyclic dependencies.
Nicolas
On 10/06/2015 04:07 PM, Maxim Uvarov wrote:
> Hello,
>
> We had patch Rujiacs to
request as much as possible
* If there are buffer left, push them to the queue.
On pktio_perf with loop interface, I see ~10% speedup with this patch
Signed-off-by: Nicolas Morey-Chaisemartin <nmo...@kalray.eu>
---
platform/linux-generic/odp_packet_io.c | 17 -
1 file chang
On 09/29/2015 04:34 PM, Bill Fischofer wrote:
> On Tue, Sep 29, 2015 at 9:15 AM, Nicolas Morey-Chaisemartin <
> nmo...@kalray.eu> wrote:
>
>
>> +
>> +/**
>> * Free packet
>> *
>> * Frees the packet into the buffer pool it was all
On 10/01/2015 05:05 PM, Bill Fischofer wrote:
> odp_packet_copy() can copy between pools, which may have different sized
> user areas, hence the size checks.
Ok. But then does it make sense to copy a subpart of the user data ?
If dst user_area >= src user_area it's going to work well but in the
On 09/29/2015 07:02 PM, Nicolas Morey-Chaisemartin wrote:
>
>
> On 09/29/2015 04:31 PM, Bill Fischofer wrote:
>>
>>
>> On Tue, Sep 29, 2015 at 9:15 AM, Nicolas Morey-Chaisemartin
>> <nmo...@kalray.eu> wrote:
>>
>> Signed-off-by: Nicolas M
201 - 300 of 594 matches
Mail list logo