[linux-yocto] [PATCH 0/2] iio: Set correct iio name

Bruce Ashfield bruce.ashfield at windriver.com
Mon May 30 21:36:41 PDT 2016


On 2016-05-30 11:38 PM, Yong Li wrote:
> Hi Bruce,
>
> The Upstream-Status:Submitted[http://www.spinics.net/lists/linux-iio/msg24331.html]
> for more information, please visit
> https://github.com/ostroproject/meta-ostro-bsp/pull/34
>
> without this patch, the device I2C name is mismatch with the IIO name:
>
> root at intel-corei7-64:/sys/bus/i2c/devices/1-0040# cat name
>
> tmp006
>
> root at intel-corei7-64:/sys/bus/i2c/devices/1-0040# cat iio\:device0/name
>
> 1-0040
>
> Sensor framework(Soletta) will use this name, the name "tmp006" is
> much better than the  "X-0040"
>

That looks fine to me .. can you resend the patch as a v2 with
that in the commit message ? I can then merge it directly from
that version.

Bruce

>
> Thanks,
> Yong
>
> 2016-05-31 2:22 GMT+08:00 Bruce Ashfield <bruce.ashfield at windriver.com>:
>> On 2016-05-28 09:50 PM, Saul Wold wrote:
>>>
>>> On Sat, 2016-05-28 at 12:34 +0800, Yong Li wrote:
>>>>
>>>> Hi Bruce Saul,
>>>>
>>>> I had submitted the patches into Kernel mail list, the concern is the
>>>> legacy application compatibility:
>>>> http://www.spinics.net/lists/linux-iio/index.html#24331
>>>>
>>>> For the Ostro OS, Using Soletta framrwork, we have tested/verified
>>>> more more than 30 different I2C
>>>> devices(https://ostroproject.org/documentation/hardware/sensors.html)
>>>> .
>>>> But only the two devices have incorrect device names(the IIO name
>>>> does
>>>> not match the I2C device name). QA team think it is a bug
>>>>
>>> So is it possible to fix the test in this case to correctly handle the
>>> legacy naming rather than make it fail?  I guess the concern here is if
>>> we "fix" the name, we will really break the legacy applications.  It's
>>> possible for those application to run in Ostro also and since Ostro is
>>> newer, it would break those rather than the other way around.
>>
>>
>> Any update on this ? I can merge the change to standard/intel, but
>> if we do that, I'd like to update the commit logs to show the upstream
>> submission, and explain that why this use case prefers to make the
>> names match (versus the upstream compatibility argument).
>>
>> That way, we'll know to carry the patch and not try to re-submit it
>> upstream later.
>>
>> Bruce
>>
>>>
>>> Sau!
>>>
>>>> Thanks,
>>>> Yong
>>>>
>>>> 2016-05-27 23:51 GMT+08:00 Saul Wold <sgw at linux.intel.com>:
>>>>>
>>>>>
>>>>> On Fri, 2016-05-27 at 10:24 -0400, Bruce Ashfield wrote:
>>>>>>
>>>>>>
>>>>>> On 2016-05-27 12:58 AM, Yong Li wrote:
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Dear Maintainers,
>>>>>>>
>>>>>>> This patch fixes the “incorrect IIO device name” issue.
>>>>>>>
>>>>>>> Please merge it into standard/base branch for linux-yocto-4.4
>>>>>>> if
>>>>>>> this looks okay.
>>>>>>
>>>>>> The change looks technically correct, just a question about if
>>>>>> these
>>>>>> are also going upstream to the mainline kernel.
>>>>>>
>>>>> Bruce,
>>>>>
>>>>> These are possibly candidates for the standard/intel branch,
>>>>> they where proposed upstream and deemed correct, but not merge-able
>>>>> due
>>>>> to creating incompatible names.
>>>>>
>>>>> Yong Li,
>>>>> What is requiring the name change in Ostro, why can't Ostro
>>>>> use the existing incorrect, but compatible name?
>>>>>
>>>>> Sau!
>>>>>>
>>>>>>
>>>>>> Bruce
>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Thanks
>>>>>>> Yong Li
>>>>>>>
>>>>>>> Yong Li (2):
>>>>>>>     iio: tmp006: Set correct iio name
>>>>>>>     iio: si7020: Set correct iio name
>>>>>>>
>>>>>>>    drivers/iio/humidity/si7020.c    | 2 +-
>>>>>>>    drivers/iio/temperature/tmp006.c | 2 +-
>>>>>>>    2 files changed, 2 insertions(+), 2 deletions(-)
>>>>>>>
>>



More information about the linux-yocto mailing list