We can discuss your proposal during the call, today.
I have a question, though. If the ODP_SHM_NONCOHERENT flag is implemeneted,
then I guess we are back to the need for a refresh() or similar function to
update the cache.
Or you could implement this flag in your implementation only, allocating
shm
On 15.01.16 11:14, Maxim Uvarov wrote:
Add test case based on previous seg fault code that pool destroy
does not fail if there are some timers in fight.
Signed-off-by: Maxim Uvarov
---
platform/linux-generic/odp_timer.c | 15 ++-
test/validation/timer/timer.c | 15
Maxim,
in general, I had no look in this test, it was working for me and I was happy.
At fast glance I have some comments. Hope you will find time to correct it.
On 15.01.16 11:14, Maxim Uvarov wrote:
Following current test logic validation test should correctly
wait for all scheduled timer expi
Hi all,
Our problem is about shmem usage so far. After some discussions we
believe we have an acceptable solution: add a new flag to
odp_shm_reserve(). So far we have ODP_SHM_SW_ONLY and ODP_SHM_PROC, we
would like to add something like ODP_SHM_NONCOHERENT or similar.
It would allow to keep th
On 01/18/2016 19:10, Ivan Khoronzhuk wrote:
+int ipc_odp_packet_sendall(odp_pktio_t pktio,
+ odp_packet_t pkt_tbl[], int num)
+{
+int ret;
+int sent = 0;
+uint64_t start_cycle;
+uint64_t diff;
+uint64_t wait;
+
+wait = odp_time_local_from_ns(1 * ODP_TIME_SEC_
merged.
Maxim.
On 01/18/2016 20:54, Ivan Khoronzhuk wrote:
ping
On 15.01.16 17:55, Stuart Haslam wrote:
On Fri, Jan 15, 2016 at 05:53:20PM +0200, Ivan Khoronzhuk wrote:
On 15.01.16 17:51, Stuart Haslam wrote:
On Wed, Jan 13, 2016 at 06:46:57PM +0200, Ivan Khoronzhuk wrote:
If packet with
On Mon, Jan 18, 2016 at 05:58:37PM +0300, Maxim Uvarov wrote:
> Stuart, can you please review this?
>
> Maxim.
I've just run it through the validation tests and all seems well, I've
have a closer look at the changes tomorrow.
--
Stuart.
___
lng-odp ma
ping
On 15.01.16 17:55, Stuart Haslam wrote:
On Fri, Jan 15, 2016 at 05:53:20PM +0200, Ivan Khoronzhuk wrote:
On 15.01.16 17:51, Stuart Haslam wrote:
On Wed, Jan 13, 2016 at 06:46:57PM +0200, Ivan Khoronzhuk wrote:
If packet with size > MTU is sent the test can crash farther,
so better to s
ping
On 13.01.16 19:18, Ivan Khoronzhuk wrote:
Now odp_time_wait_ns() can be used.
Signed-off-by: Ivan Khoronzhuk
---
test/validation/pktio/pktio.c | 13 +
1 file changed, 1 insertion(+), 12 deletions(-)
diff --git a/test/validation/pktio/pktio.c b/test/validation/pktio/pktio.c
ping
On 13.01.16 19:38, Ivan Khoronzhuk wrote:
The odp_pktio_stop() called is supposed to return error.
Signed-off-by: Ivan Khoronzhuk
---
test/validation/pktio/pktio.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/test/validation/pktio/pktio.c b/test/validation/pktio/
ping
On 14.01.16 12:56, Ivan Khoronzhuk wrote:
In case of one loop pktio the MAC address is not set in the packets
but should be.
Signed-off-by: Ivan Khoronzhuk
---
test/validation/pktio/pktio.c | 18 ++
1 file changed, 14 insertions(+), 4 deletions(-)
diff --git a/test/val
Old version of OpenSSL require SET_TAG to be called before
decrypting the data. New versions are compatible either way
Signed-off-by: Nicolas Morey-Chaisemartin
---
Repost due to previous mail tagged as SPAM
platform/linux-generic/odp_crypto.c | 3 +--
1 file changed, 1 insertion(+), 2 deleti
On 18.01.16 11:59, Maxim Uvarov wrote:
On 01/15/2016 15:57, Stuart Haslam wrote:
On Mon, Dec 21, 2015 at 01:56:11PM +0300, Maxim Uvarov wrote:
2 example ipc pktio applications create ipc pktio to each other and do
packet transfer, validation magic numbers and packets sequence counters
inside
On 18 January 2016 at 16:25, Ola Liljedahl wrote:
> On 18 January 2016 at 14:25, Petri Savolainen
> wrote:
>
>> Added validation test for exchange operations.
>>
>> Signed-off-by: Petri Savolainen
>> ---
>> test/validation/synchronizers/synchronizers.c | 54
>> +++
>> t
Merged.
On 01/18/2016 16:40, Savolainen, Petri (Nokia - FI/Espoo) wrote:
Not sure if "odp_cpu_cycles_*" will cause problems in git log history. May be Maxim can
remove the '*' from the string during merge (e.g. just "... cpu_cycles calls").
done.
Reviewed-by: Petri Savolainen
-Origi
On 18 January 2016 at 14:25, Petri Savolainen
wrote:
> Added validation test for exchange operations.
>
> Signed-off-by: Petri Savolainen
> ---
> test/validation/synchronizers/synchronizers.c | 54
> +++
> test/validation/synchronizers/synchronizers.h | 1 +
> 2 files c
Stuart, can you please review this?
Maxim.
On 01/11/2016 14:39, Petri Savolainen wrote:
This patch set adds support for multi-queue pktio for poll mode queues.
v3:
* Rebased
v2:
* Bug correction in patch 2/6: bad netmap initialization when using
the old API.
Matias Elo (1):
lin
strange that on make check l2fwd test fails due to small packet rate
1405 maximum packets per second.
without patch:
6475 maximum packets per second.
test consumes that less then 5000 is error.
Please check on your machine.
Maxim.
On 01/15/2016 16:27, Bill Fischofer wrote:
On Thu, Jan 14
Not sure if "odp_cpu_cycles_*" will cause problems in git log history. May be
Maxim can remove the '*' from the string during merge (e.g. just "...
cpu_cycles calls").
Reviewed-by: Petri Savolainen
> -Original Message-
> From: EXT Ivan Khoronzhuk [mailto:ivan.khoronz...@linaro.org]
>
Use full memory barrier from API instead of the internal
function.
Signed-off-by: Petri Savolainen
---
platform/linux-generic/include/odp_atomic_internal.h | 6 --
platform/linux-generic/odp_barrier.c | 8
2 files changed, 4 insertions(+), 10 deletions(-)
diff --git
Replace internal non-relaxed atomics with API calls. Timer and
spinlock were not converted since those use flags and 128 variables.
Signed-off-by: Petri Savolainen
---
platform/linux-generic/include/odp_pool_internal.h | 1 -
platform/linux-generic/odp_pool.c | 1 -
platform/l
Added validation test for all operations with non-relaxed
memory model. Mainly testing that functions are available.
Signed-off-by: Petri Savolainen
---
test/validation/synchronizers/synchronizers.c | 82 +++
test/validation/synchronizers/synchronizers.h | 1 +
2 files c
Added validation test for exchange operations.
Signed-off-by: Petri Savolainen
---
test/validation/synchronizers/synchronizers.c | 54 +++
test/validation/synchronizers/synchronizers.h | 1 +
2 files changed, 55 insertions(+)
diff --git a/test/validation/synchronizers/s
Added validation test for CAS operations. Maintained the structure
of other atomic tests: one function increments and another decrements
a variable.
Signed-off-by: Petri Savolainen
---
test/validation/synchronizers/synchronizers.c | 83 +++
test/validation/synchronizers/s
Added validation tests for atomic max and min operations.
Results validation is a simple compare, since absolute min/max
values depend on the number of threads.
Signed-off-by: Petri Savolainen
---
test/validation/synchronizers/synchronizers.c | 111 --
test/validation/syn
Added 32 and 64 bit atomic exchange operations in relaxed
memory ordering. These can be used e.g. to initialize a counter
while counting.
Signed-off-by: Petri Savolainen
---
include/odp/api/atomic.h | 27 +++
1 file changed, 27 insertions(+)
diff --git a/include/odp/api/
Implemented 32 and 64 bit atomic exchange operations.
Signed-off-by: Petri Savolainen
---
platform/linux-generic/include/odp/atomic.h | 16
1 file changed, 16 insertions(+)
diff --git a/platform/linux-generic/include/odp/atomic.h
b/platform/linux-generic/include/odp/atomic.h
i
Implemented 64 bit non-relaxed atomic operations. Additional
ordering is not needed in case of using ATOMIC_OP since it
includes both acquire and release ordering.
Signed-off-by: Petri Savolainen
---
platform/linux-generic/include/odp/atomic.h | 83 +
1 file changed,
Added 64 bit versions of the same set of non-relaxed atomic
operations that are already defined for 32 bit.
Signed-off-by: Petri Savolainen
---
include/odp/api/atomic.h | 99
1 file changed, 99 insertions(+)
diff --git a/include/odp/api/atomic.h
Added RELEASE and ACQUIRE-and-RELEASE versions of 32 bit CAS.
These are commonly needed operations in lock-free algorithms.
Signed-off-by: Petri Savolainen
---
include/odp/api/atomic.h | 86 +---
1 file changed, 67 insertions(+), 19 deletions(-)
diff
Implemented 32bit cas_rel and cas_acq_rel calls.
Signed-off-by: Petri Savolainen
---
platform/linux-generic/include/odp/atomic.h | 37 ++---
1 file changed, 28 insertions(+), 9 deletions(-)
diff --git a/platform/linux-generic/include/odp/atomic.h
b/platform/linux-generi
Rename release ordering from _rls_ to _rel_ since it matches
better with _acq_rel_, which is going to be used for C11
memory_order_acq_rel order.
Signed-off-by: Petri Savolainen
---
include/odp/api/atomic.h| 8
platform/linux-generic/include/odp/atomic.h | 6 +++---
Completed additions to atomic API. Added exchange operation in relaxed memory
model. Added CAS operations in rel and acq_rel memory order. Added 64 bit
versions of all non-relaxed operations defined so far. Added all missing
validation tests. Replaced internal atomics usage in lock, etc code
(a
v6 is sent.
On 18.01.16 14:58, Savolainen, Petri (Nokia - FI/Espoo) wrote:
-Original Message-
From: EXT Ivan Khoronzhuk [mailto:ivan.khoronz...@linaro.org]
Sent: Monday, January 18, 2016 2:24 PM
To: Savolainen, Petri (Nokia - FI/Espoo); lng-odp@lists.linaro.org
Subject: Re: [lng-odp]
https://bugs.linaro.org/show_bug.cgi?id=1906
Signed-off-by: Ivan Khoronzhuk
---
Since v5:
- add res instead of "1" while wrap
- get cycle stamps a multiple of res instead direct valuues.
- move resolution test before diff test
Since v3:
- modified log "wrap is not detected"
- increased try num f
Merged,
Maxim.
On 01/18/2016 15:22, Bala Manoharan wrote:
For the series:
Reviewed-by: Balasubramanian Manoharan
Regards,
Bala
On 15 January 2016 at 22:58, Ivan Khoronzhuk wrote:
This patch series corrects classification tests to be a little bit
adoptive.
Since v1:
- removed patch for dele
> -Original Message-
> From: EXT Ivan Khoronzhuk [mailto:ivan.khoronz...@linaro.org]
> Sent: Monday, January 18, 2016 2:24 PM
> To: Savolainen, Petri (Nokia - FI/Espoo); lng-odp@lists.linaro.org
> Subject: Re: [lng-odp] [PATCH v5] validation: system: add validation tests
> for odp_cpu_cyc
please run ./scripts/checkpatch.pl
on patches, some of them have errors.
Maxim.
On 01/18/2016 13:09, hongbo.zh...@linaro.org wrote:
From: Hongbo Zhang
v4->v5 changes:
- rebase to latest api-next branch
v3->v4 changes:
- update patch 9 of platform naming
- add new patch 12 to rename cpu
On 8 January 2016 at 14:41, Petri Savolainen
wrote:
> Completed additions to atomic API. Added exchange operation in relaxed
> memory
> model. Added CAS operations in rel and acq_rel memory order. Added 64 bit
> versions of all non-relaxed operations defined so far. Added all missing
> validation
WARNING: line over 80 characters
#73: FILE: platform/linux-generic/arch/x86/odp_sysinfo_parse.c:27:
+sysinfo->cpu_hz[id] = (uint64_t)(ghz * 10.0);
On 01/18/2016 13:10, hongbo.zh...@linaro.org wrote:
From: Hongbo Zhang
This patch moves the x86 system info codes into the
On 18.01.16 13:37, Savolainen, Petri (Nokia - FI/Espoo) wrote:
-Original Message-
From: EXT Ivan Khoronzhuk [mailto:ivan.khoronz...@linaro.org]
Sent: Friday, January 15, 2016 5:07 PM
To: lng-odp@lists.linaro.org
Cc: Savolainen, Petri (Nokia - FI/Espoo); Ivan Khoronzhuk
Subject: [lng-
For the series:
Reviewed-by: Balasubramanian Manoharan
Regards,
Bala
On 15 January 2016 at 22:58, Ivan Khoronzhuk wrote:
> This patch series corrects classification tests to be a little bit
> adoptive.
>
> Since v1:
> - removed patch for deleting input queue, as it should be discussed first
> -
Merged.
Maxim.
On 01/18/2016 12:04, Savolainen, Petri (Nokia - FI/Espoo) wrote:
Reviewed-by: Petri Savolainen
-Original Message-
From: lng-odp [mailto:lng-odp-boun...@lists.linaro.org] On Behalf Of EXT
Stuart Haslam
Sent: Friday, January 15, 2016 5:18 PM
To: lng-odp@lists.linaro.org
> -Original Message-
> From: EXT Ivan Khoronzhuk [mailto:ivan.khoronz...@linaro.org]
> Sent: Friday, January 15, 2016 5:07 PM
> To: lng-odp@lists.linaro.org
> Cc: Savolainen, Petri (Nokia - FI/Espoo); Ivan Khoronzhuk
> Subject: [lng-odp] [PATCH v5] validation: system: add validation tests
From: Hongbo Zhang
The legacy variable cpu_hz holds the max frequency for odp_cpu_hz_max(),
so rename it to cpu_hz_max for better meaning.
Signed-off-by: Hongbo Zhang
---
platform/linux-generic/arch/mips64/odp_sysinfo_parse.c | 2 +-
platform/linux-generic/arch/powerpc/odp_sysinfo_parse.c | 2
From: Hongbo Zhang
When per-CPU framework was introduced, it was only implemented on x86,
for other platforms, only the model_str[0] and cpu_hz[0] are set to pass
compile, this patch set all values for model_str[] and cpu_hz[] on the
PowerPC platform.
Signed-off-by: Hongbo Zhang
---
.../linux-
From: Hongbo Zhang
When per-CPU framework was introduced, it was only implemented on x86,
for other platforms, only the model_str[0] and cpu_hz[0] are set to pass
compile, this patch set all values for model_str[] and cpu_hz[] on the
MIPS platform.
Signed-off-by: Hongbo Zhang
---
.../linux-gen
From: Hongbo Zhang
The values of cpu_arch_str were deleted due to codes reorganization, this
patch adds values to them again.
Signed-off-by: Hongbo Zhang
---
platform/linux-generic/arch/mips64/odp_sysinfo_parse.c | 1 +
platform/linux-generic/arch/powerpc/odp_sysinfo_parse.c | 1 +
platform/l
From: Hongbo Zhang
All the CPU frequency API return value type is uint64_t, this patch makes
them return 0 on failure instead of -1.
Signed-off-by: Hongbo Zhang
---
include/odp/api/cpu.h | 4
platform/linux-generic/arch/linux/odp_sysinfo_parse.c | 2 +-
From: Hongbo Zhang
This patch moves the PowerPC system info codes into the newly added PowerPC
specific platform file.
This patch also creates syslink to arch/linux/odp_cpu_cycles.c for PowerPC.
Signed-off-by: Hongbo Zhang
---
configure.ac | 1 +
platform
From: Hongbo Zhang
This patch moves the ARM system info codes into the general arch file,
ARM sysinfo isn't implemented now and there is no need to create specific
arch file for it, but this can serve as the default implementaton instead.
Signed-off-by: Hongbo Zhang
---
platform/linux-generic/
From: Hongbo Zhang
This patch moves the MIPS system info codes into the newly added MIPS
specific platform file.
Signed-off-by: Hongbo Zhang
---
platform/linux-generic/Makefile.am | 1 +
.../linux-generic/arch/mips64/odp_sysinfo_parse.c | 53 ++
platform/l
From: Hongbo Zhang
This function will be moved to their own platform specific files too,
it is better to be named with a odp_ prefix.
Signed-off-by: Hongbo Zhang
---
platform/linux-generic/odp_system_info.c | 12 ++--
1 file changed, 6 insertions(+), 6 deletions(-)
diff --git a/platfo
From: Hongbo Zhang
This patch moves the x86 system info codes into the newly added x86
specific platform file.
Signed-off-by: Hongbo Zhang
---
platform/linux-generic/Makefile.am | 6 +-
.../linux-generic/arch/x86/odp_sysinfo_parse.c | 70 ++
platform/li
From: Hongbo Zhang
It is common way that the general top layer calls a uniform interface to
initialize data structrure for different plarform, this patch introduces
odp_sysinfo_parser() instead of using different interfaces for each
platform, and also removes using structure odp_compiler_info_t s
From: Hongbo Zhang
This is a pre-step for separating platform specific codes into their own
arch files.
Value of cpu_arch_str will be set later in separate arch parser code.
Signed-off-by: Hongbo Zhang
---
platform/linux-generic/include/odp_internal.h | 1 +
platform/linux-generic/odp_system_i
From: Hongbo Zhang
v4->v5 changes:
- rebase to latest api-next branch
v3->v4 changes:
- update patch 9 of platform naming
- add new patch 12 to rename cpu_hz to cpu_hz_max
v2->v3 changes:
use "api: cpu:" tag in patch 8/11 title instead of "linux-generic: sysinfo"
v1->v2 changes:
- don't cr
On Mon, Jan 18, 2016 at 12:59:44PM +0300, Maxim Uvarov wrote:
> On 01/15/2016 15:57, Stuart Haslam wrote:
> >On Mon, Dec 21, 2015 at 01:56:11PM +0300, Maxim Uvarov wrote:
> >>2 example ipc pktio applications create ipc pktio to each other and do
> >>packet transfer, validation magic numbers and pac
Merged,
Maxim.
On 01/15/2016 19:05, Ivan Khoronzhuk wrote:
On 14.01.16 23:21, Bill Fischofer wrote:
Restructure odp_init_global() and odp_init_local() so that they recover
cleanly if initialization fails. At exit any partial initialization is
reversed so that system is in the same state as be
On 01/15/2016 15:57, Stuart Haslam wrote:
On Mon, Dec 21, 2015 at 01:56:11PM +0300, Maxim Uvarov wrote:
2 example ipc pktio applications create ipc pktio to each other and do
packet transfer, validation magic numbers and packets sequence counters
inside it.
It looks like there's a race somewhe
On 01/15/2016 16:01, Stuart Haslam wrote:
On Mon, Dec 21, 2015 at 01:56:10PM +0300, Maxim Uvarov wrote:
Signed-off-by: Maxim Uvarov
---
platform/linux-generic/Makefile.am | 3 +
.../linux-generic/include/odp_buffer_internal.h| 3 +
.../linux-generic/include/odp_packe
This seems to be a common opinion. I think it is nevertheless fair to give
a chance to Nicolas to explain his case.
On 18 January 2016 at 10:02, Ola Liljedahl wrote:
> On 15 January 2016 at 15:48, Christophe Milard <
> christophe.mil...@linaro.org> wrote:
>
>> On Tuesday's ARCH call, we will be
Reviewed-by: Petri Savolainen
> -Original Message-
> From: lng-odp [mailto:lng-odp-boun...@lists.linaro.org] On Behalf Of EXT
> Stuart Haslam
> Sent: Friday, January 15, 2016 5:18 PM
> To: lng-odp@lists.linaro.org
> Subject: [lng-odp] [PATCH API-NEXT] api: pktio: remove unused
> ODP_PKTI
On 15 January 2016 at 15:48, Christophe Milard wrote:
> On Tuesday's ARCH call, we will be discussing the above topic.
> Nicolas (Kalray) will try to join, but he is very uncertain at this stage
> (if he can be there)...
>
Cache coherence is the norm today, non-coherent is an exception.
My opini
64 matches
Mail list logo