Hi all,
As some of you know we at ARM have been working on prototyping a packet
processing application using ODP as the HW abstraction API one of the
goals of this work is to ³validate² the API from the application point of
view and provide contributions where it makes sense if we think that it
especially when trying to
minimize traffic loss.
I don't think resource creation and tearing down is dataplane job, these
functions should be called from a control plane thread.
Thanks,
Alex
On 17 August 2015 at 11:59, Mario Torrecillas Rodriguez
mario.torrecillasrodrig...@arm.com wrote:
Hi all
The problem with using count as a way of determining whether the slot is
unallocated or not is that count keeps the current number of threads in the
group. If a group is created with no threads in it (which is allowed by the
implementation as far as I can see), then no thread will be able to
On a related noteŠ
It is my current understanding that, with the current set of APIs,
classification has to be performed right after IO. What makes me think
that is that the classification API seems to be linked to the packetIO API.
So, my question is, firstly, am I right with my assumption? And
Added odp_errno.c and errno.h
Changed odp_packet_io and odp_timer to use it.
Signed-off-by: Mario Torrecillas Rodriguez mario.torrecillasrodrig...@arm.com
---
(This code contribution is provided under the terms of agreement LES-LTM-21309)
Changes from previous version:
* Fixed pktio test issue
Added odp_errno.c and odp_errno.h
Changed odp_packet_io and odp_timer to use it.
Signed-off-by: Mario Torrecillas Rodriguez mario.torrecillasrodrig...@arm.com
---
(This code contribution is provided under the terms of agreement LES-LTM-21309)
Changes from previous version:
* Renamed odp_errno.h
Added odp_errno.c and odp_errno.h
Changed odp_packet_io and odp_timer to use it.
Signed-off-by: Mario Torrecillas Rodriguez mario.torrecillasrodrig...@arm.com
---
(This code contribution is provided under the terms of agreement LES-LTM-21309)
Changes from previous version:
* Added doxygen group
Added odp_errno.c and odp_errno.h
Changed odp_packet_io and odp_timer to use it.
Signed-off-by: Mario Torrecillas Rodriguez mario.torrecillasrodrig...@arm.com
---
(This code contribution is provided under the terms of agreement LES-LTM-21309)
Changes from previous version:
* Changed printf
Added odp_errno.c and odp_errno.h
Changed odp_packet_io and odp_timer to use it.
Signed-off-by: Mario Torrecillas Rodriguez mario.torrecillasrodrig...@arm.com
---
(This code contribution is provided under the terms of agreement LES-LTM-21309)
Changes from previous version:
* Changed doxygen
Thanks for the review Petri. A short comment inlined.
On 20/01/2015 13:38, Savolainen, Petri (NSN - FI/Espoo)
petri.savolai...@nsn.com wrote:
--- /dev/null
+++ b/platform/linux-generic/include/api/odp_errno.h
@@ -0,0 +1,61 @@
+/* Copyright (c) 2015, Linaro Limited
+ * All rights reserved.
Added odp_errno.c and odp_errno.h
Changed odp_packet_io and odp_timer to use it.
Signed-off-by: Mario Torrecillas Rodriguez mario.torrecillasrodrig...@arm.com
---
(This code contribution is provided under the terms of agreement LES-LTM-21309)
Changes from previous version:
* Moved __odp_errno
Added odp_errno.c and odp_errno.h
Changed odp_packet_io and odp_timer to use it.
Signed-off-by: Mario Torrecillas Rodriguez mario.torrecillasrodrig...@arm.com
---
(This code contribution is provided under the terms of agreement LES-LTM-21309)
platform/linux-generic/Makefile.am | 2
The barrier tests seems to be fine actually, it always either hangs on the
no_barrier test, or doesn¹t hang at all. I¹m looking into this now, it might
be a design problem rather than a bug in the code.
Mario.
From: Mike Holmes mike.hol...@linaro.org
Date: Friday, 9 January 2015 18:29
To:
barrier was being
used.
Mario.
On 12/01/2015 11:25, Mario Torrecillas Rodriguez
mario.torrecillasrodrig...@arm.com wrote:
First set of synchronizer tests. This patch includes
tests for locks, barriers and atomics.
Signed-off-by: Mario Torrecillas Rodriguez
mario.torrecillasrodrig...@arm.com
First set of synchronizer tests. This patch includes
tests for locks, barriers and atomics.
Signed-off-by: Mario Torrecillas Rodriguez mario.torrecillasrodrig...@arm.com
---
(This code contribution is provided under the terms of agreement LES-LTM-21309)
Changes from previous version:
Fixed issue
First set of synchronizer tests. This patch includes
tests for locks, barriers and atomics.
Signed-off-by: Mario Torrecillas Rodriguez mario.torrecillasrodrig...@arm.com
---
(This code contribution is provided under the terms of agreement LES-LTM-21309)
Changes from previous version:
Addressed
First set of synchronizer tests. This patch includes tests for locks,
barriers and atomics (sunny day tests only).
Signed-off-by: Mario Torrecillas Rodriguez mario.torrecillasrodrig...@arm.com
---
(This code contribution is provided under the terms of agreement LES-LTM-21309)
Changes from
too (e.g. now gitignore contains the
binary).
Mario.
On 18/12/2014 10:51, Mario Torrecillas Rodriguez
mario.torrecillasrodrig...@arm.com wrote:
First set of synchronizer tests. This patch includes tests for locks,
barriers and atomics (sunny day tests only).
Signed-off-by: Mario Torrecillas
First set of synchronizer tests. This patch includes tests for locks,
barriers and atomics (sunny day tests only).
Most of this code comes from Barry Spinney and Yan Songming.
Please note that this code depends on ticketlock_trylock
Signed-off-by: Mario Torrecillas Rodriguez
First set of synchronizer tests. This patch includes tests for locks,
barriers and atomics (sunny day tests only).
Most of this code comes from Barry Spinney and Yan Songming.
Please note that this code depends on ticketlock_trylock
Signed-off-by: Mario Torrecillas Rodriguez
This version of the patch should fix all the issues mentioned so far (by
Jerin and Maxim, I believe).
Cheers,
Mario.
On 17/12/2014 16:31, Mario Torrecillas Rodriguez
mario.torrecillasrodrig...@arm.com wrote:
First set of synchronizer tests. This patch includes tests for locks,
barriers
on ticketlock_trylock
Signed-off-by: Mario Torrecillas Rodriguez mario.torrecillasrodrig...@arm.com
---
test/validation/Makefile.am |7 +-
test/validation/common/odp_common.c | 241 ++
test/validation/common/odp_common.h | 28 +
test/validation/common
.
Mario.
On 16/12/2014 09:44, Mario Torrecillas Rodriguez
mario.torrecillasrodrig...@arm.com wrote:
(This code contribution is provided under the terms of agreement
LES-LTM-21309)
First set of synchronizer tests. This patch includes tests for locks,
barriers and atomics (sunny day tests only).
Most
I have combined both patches into a single one using Barry¹s as the
baseline. I need to fix a couple of minor issues that I have and do some
cleanup but I¹ll most likely send the patch tomorrow so that we can continue
working on this unified set of synchroniser tests from now on.
Like Mike said,
I don¹t think cur_ticket will be read/written by more than one thread
simultaneously, only the one which owns the lock will attempt to update it.
Mario.
On 03/12/2014 17:16, Maxim Uvarov maxim.uva...@linaro.org wrote:
On 12/03/2014 05:20 PM, Ola Liljedahl wrote:
Ping!
Needed more review for
Hi Petri, all,
According to the document containing the list of remaining tasks, these two
APIs need to be made internal (at least, this is explicitly mentioned in the
case of odp_compiler. For odp_byteorder it is not clear whether the plan is to
move it to its own header file or merge it with
26 matches
Mail list logo