Hi Maxim.
I cannot reproduce that, and your trace seems to indicate the
fallback to normal pages does work.
I do see the test failure in you log, though, but cannot relate it to
ishm (or to anything else). Can you provide me with more log/ help me
debug?
Christophe
My run seem to work:
Distribu
On 01/20 16:11:56, Petri Savolainen wrote:
> Used event API as the first example of an ABI spec file. Used
> the same default architecture file initially for all
> architectures. Default ABI files avoid multiple copies
> of the same definition.
Please consider an alternative approach to achieving
On 23 January 2017 at 13:45, Bill Fischofer wrote:
> Moving this discussion on the ODP mailing list rather than the
> Internal list as that way it will be archived.
>
> The existing ODP controls over scheduling include schedule groups as
> well as queue priorities. The former is a strict requireme
On Mon, Jan 23, 2017 at 9:38 AM, Christophe Milard
wrote:
> _ishm prereserves address space for _ODP_ISHM_SINGLE_VA allocated memory
> by reserving unreachable "memory", using PROT_NONE, hence really allocating
> virtual space (as the memory cannot be hit).
> The problem came when trying to use so
OK. For the v2 series:
Reviewed-and-tested-by: Bill Fischofer
On Mon, Jan 23, 2017 at 5:09 AM, Savolainen, Petri (Nokia - FI/Espoo)
wrote:
>> > +#ifndef ODP_ABI_EVENT_H_
>> > +#define ODP_ABI_EVENT_H_
>> > +
>> > +#ifdef __cplusplus
>> > +extern "C" {
>> > +#endif
>> > +
>> > +#include
>> > +
Reviewed-and-tested-by: Brian Brooks
On 01/23 14:46:04, Mike Holmes wrote:
> Greatly reduce the proliferation of helper includes that every app needs
> Make the public helper API very obvious
> Fix recent inclusion of table APIs that were not in the helper include dir and
> were not exported duri
Clean up all the includes and replace them with odph_api.h
Signed-off-by: Mike Holmes
---
test/common_plat/common/odp_cunit_common.c | 2 +-
test/common_plat/performance/odp_crypto.c | 2 +-
test/common_plat/performance/odp_l2fwd.c
Adding the previously missing table functions to the public helper api
allows the tests to use just odph_api.h
Signed-off-by: Mike Holmes
---
helper/Makefile.am | 10 +-
helper/cuckootable.c | 2 +-
helper/hashtable.c
clean up all the examples and just use odph_api.h
Signed-off-by: Mike Holmes
---
example/classifier/odp_classifier.c | 4 +---
example/generator/odp_generator.c | 6 +-
example/ipsec/odp_ipsec.c | 6 +-
example/ipsec/odp_ipsec_fwd_db.h| 1 -
example/ipsec/
Rather than have all applications include a variable number of helper
api files, consolidate them into one include odph_api.h
Signed-off-by: Mike Holmes
---
helper/Makefile.am | 1 +
helper/include/odp/helper/odph_api.h | 35 +++
helper/test/chk
Greatly reduce the proliferation of helper includes that every app needs
Make the public helper API very obvious
Fix recent inclusion of table APIs that were not in the helper include dir and
were not exported during install.
Mike Holmes (4):
helper: add odph_api.h for existing exported headers
Merged,
Maxim.
On 01/20/17 18:31, Bill Fischofer wrote:
> For the v2 series:
>
> Reviewed-and-tested-by: Bill Fischofer
>
> On Fri, Jan 20, 2017 at 3:14 AM, Petri Savolainen
> wrote:
>> Implement round up to power of two as a macro. Macro
>> enables preprocessor to calculate the constant value
Moving this discussion on the ODP mailing list rather than the
Internal list as that way it will be archived.
The existing ODP controls over scheduling include schedule groups as
well as queue priorities. The former is a strict requirement (threads
can only receive events from queues that belong t
On 01/23/17 11:47, Christophe Milard wrote:
> Fixing failure due to lack of huge pages.
> Fixes: https://bugs.linaro.org/show_bug.cgi?id=2842
>
> Signed-off-by: Christophe Milard
> ---
> platform/linux-generic/_ishm.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/
Merged,
Maxim.
On 01/23/17 17:35, Matias Elo wrote:
> The application calls each packet function TEST_REPEAT_COUNT times and
> calculates the average cycle count per call.
>
> Main options:
> -b: Burst size for *_multi operations
> -i: Benchmark index to run indefinitely
>
> The application prin
> -Original Message-
> From: Mike Holmes [mailto:mike.hol...@linaro.org]
> Sent: Monday, January 23, 2017 3:01 PM
> To: Savolainen, Petri (Nokia - FI/Espoo) labs.com>
> Cc: lng-odp@lists.linaro.org
> Subject: Re: [lng-odp] [PATCH v4 3/4] helper: move thread implementation
> under platfor
The application calls each packet function TEST_REPEAT_COUNT times and
calculates the average cycle count per call.
Main options:
-b: Burst size for *_multi operations
-i: Benchmark index to run indefinitely
The application prints benchmark results as a table, which can be copied
as is to a sprea
_ishm prereserves address space for _ODP_ISHM_SINGLE_VA allocated memory
by reserving unreachable "memory", using PROT_NONE, hence really allocating
virtual space (as the memory cannot be hit).
The problem came when trying to use some of this preallocated space:
strangely, if a new mapping on the p
> On 19 Jan 2017, at 17:39, Maxim Uvarov wrote:
>
> On 01/14/17 19:20, Bill Fischofer wrote:
>> --- a/test/common_plat/performance/Makefile.am
>>> +++ b/test/common_plat/performance/Makefile.am
>>> @@ -2,9 +2,12 @@ include $(top_srcdir)/test/Makefile.inc
>>>
>>> TESTS_ENVIRONMENT += TEST_DIR=${
On 23 January 2017 at 04:06, Savolainen, Petri (Nokia - FI/Espoo)
wrote:
>
>
>
>> --- /dev/null
>> +++ b/helper/include/odp/helper/platform/linux-generic/threads_extn.h
>
>
> This file and these functions are not odp-linux specific. Those depend mainly
> on Linux and somewhat on ODP, but should n
> > +#ifndef ODP_ABI_EVENT_H_
> > +#define ODP_ABI_EVENT_H_
> > +
> > +#ifdef __cplusplus
> > +extern "C" {
> > +#endif
> > +
> > +#include
> > +
> > +/** @internal Dummy type for strong typing */
> > +typedef struct { char dummy; /**< @internal Dummy */ }
> _odp_abi_event_t;
>
> Why change from
The kernel has used a set of spinlocks to help dealing with atomics in
certain conditions (architecture, length of bitfield...).
A discussion on what atomi operations are needed may be of value to ODP too:
https://lwn.net/Articles/695257/
On 23 January 2017 at 10:49, Joe Savage wrote:
> > Any a
> Any atomics can be emulated in SW (using compiler built-ins or locks
> directly). The point here is the missing HW support:
> * E.g. MIPS, Power, ARMv7 do not have 128 bit CAS
> * 128 bit fetch-and-add is not supported in any of the architectures
>
> We need to ensure on any operations added
> --- /dev/null
> +++ b/helper/include/odp/helper/platform/linux-generic/threads_extn.h
This file and these functions are not odp-linux specific. Those depend mainly
on Linux and somewhat on ODP, but should not depend at all on the odp-linux
implementation.
Should it be moved to here:
/help
Bug switched to resolved state with corresponding git hash id for commit or
number of commits.
I will run some tests in arm vm which should catch this.
Maxim.
On 23 January 2017 at 11:47, Christophe Milard wrote:
> Fixing failure due to lack of huge pages.
> Fixes: https://bugs.linaro.org/show
> -Original Message-
> From: Brian Brooks [mailto:brian.bro...@linaro.org]
> Sent: Friday, January 20, 2017 7:47 PM
> To: Francois Ozog
> Cc: Bill Fischofer ; Joe Savage
> ; Maxim Uvarov ; Savolainen,
> Petri (Nokia - FI/Espoo) ; lng-
> o...@lists.linaro.org; nd
> Subject: Re: [lng-odp]
26 matches
Mail list logo