[OE-core] [PATCH 1/5] rootfs.py: two changes regarding log checking

ChenQi Qi.Chen at windriver.com
Thu Mar 19 02:16:02 UTC 2015


On 03/18/2015 11:46 PM, Aníbal Limón wrote:
> Hi Paul, Chen,
>
> Days ago i sent a patchset for handle INCOMPATIBLE_LICENSE in manifest 
> creation [1] that contains
> a rewrite of license_manifest_creation from shell to python, i mean 
> now these warnings appears.
>
> Also i sent a similar patches like Chen's ones to avoid warnings 
> [2][3],  You said (Paul) that packagegroups
> should not have files, how we can address these case to avoid license 
> warnings in packagegroups?
>

Hi Alimon,

Paul suggested to skip checking if the package contains no file, because 
in such case the license is not relevant as long as the final image is 
concerned.
I will send out a new patchset using the above approach.

Best Regards,
Chen Qi

> Cheers,
>     alimon
>
> [1] 
> http://lists.openembedded.org/pipermail/openembedded-core/2015-March/102684.html
> [2] 
> http://lists.openembedded.org/pipermail/openembedded-core/2015-March/102683.html
> [3] 
> http://lists.openembedded.org/pipermail/openembedded-core/2015-March/102682.html
>
> On 18/03/15 03:48, Paul Eggleton wrote:
>> On Wednesday 18 March 2015 17:16:10 Chen Qi wrote:
>>> 1. Extend the regular expression to also catch '^WRARNING:' in
>>> _log_check_warn. Warnings from bb.note or bbnote begin with 
>>> 'WARNING:'. So
>>> if we decide to catch warnings at rootfs time, we should not ignore 
>>> those
>>> produced by the build system itself.
>> FYI, I still have an unfinished patchset to ensure bbwarn calls from 
>> shell
>> functions get piped through to BitBake's event system, with the 
>> result that
>> such warnings actually get logged and displayed by the BitBake UI. (That
>> doesn't mean this patch shouldn't go in in the mean time, though.)
>>
>> Cheers,
>> Paul
>>
>
>
>




More information about the Openembedded-core mailing list