kicad-developers team mailing list archive

[Dick Hollenbeck <dick@xxxxxxxxxxx>]

On 01/23/2013 10:09 AM, Dick Hollenbeck wrote:
> On 01/23/2013 08:02 AM, Miguel Angel Ajo Pelayo wrote:
>> Somebody told me about this, I think it was Brian.
>>
>> I think we must have introduced some change to the cmake scripts that change the build
>> order in this case, or my cmake scripts were not good... It was my first cmake script.
>>
>> This step should happen after swig executed already (and not before), else the pcbnew.py
>> won't exist.
>
> I think line 446 in pcbnew/CMakeLists.txt is suspicious.
>
> Usually use the target name, without path, in the DEPENDS lines.
>
>
> Miguel, if you want me to look at this I will, but I need to know how to reproduce it. 
> Perhaps the
> reproduce-ability is depending on the
>
>
> make -j<n>
>
> setting, i.e. <n>.


This is true.

-j1 works

-j5 results in the reported failure.

So it could be a CMake or make bug also, rather than a problem with pcbnew/CMakeLists.txt