[CMake] Random behavior of CTEST_USE_LAUNCHERS

[Nils Gladitz]

This seems to happen with my windows builds (icl 11.1, 12.1 and 
microsoft cl from VS 2005/2010 cmake 2.8.7) in some shape or form but my 
gcc and clang builds on linux are warning free (cmake 2.8.9) though the 
clang build without launchers actually had some warnings in it.

I exclude a 3rdparty subtree of the project by adding its name to 
CTEST_CUSTOM_WARNING_EXCEPTION.
With e.g. intel 11.1 (+ VS 2005) most warnings from that subtree are 
ignored (many #1786 CRT strcpy etc. deprecation warnings) but the 
warnings in one of those files survives even though it looks identical 
(same subtree, same #1786 warning) to the ones that were ignored.

Very puzzling.

Nils

On 10/04/2012 11:16 AM, David Cole wrote:
> On Thu, Oct 4, 2012 at 4:36 AM, Nils Gladitz <gladitz at sci-vis.de> wrote:
>> Hm ... yes weirdly enough it does look like some warnings (most) are still
>> suppressed but not all.
>>
> I've seen this happen when warning text bleeds across multiple lines,
> and both lines match some warning expression.
>
> Especially with clang which has a "warning" summary line at the end of
> a compile that had warnings...
>
>
>> Nils
>>
>>
>>
>> On 10/04/2012 09:52 AM, Julien Malik wrote:
>>> Hi Nils,
>>>
>>> I do have a bunch of CTEST_CUSTOM_WARNING_EXCEPTION in my project too, and
>>> they are treated correctly, with or without launchers.
>>> I also only read them with
>>> "ctest_read_custom_files(${CTEST_BINARY_DIRECTORY})" just after the
>>> ctest_configure step, the actual CTestCustom.cmake file being also generated
>>> at configure time.
>>>
>>> Julien
>>>
>>>
>>> Le 04/10/2012 09:14, Nils Gladitz a écrit :
>>>> I noticed I made the same mistake (I forgot to set it in the ctest
>>>> script).
>>>> Now that launchers work my CTEST_CUSTOM_WARNING_EXCEPTION seems to have
>>>> stopped working though.
>>>>
>>>> Do those work differently with launchers?
>>>>
>>>> Do I have to set them in two places as well perhaps?
>>>> Currently I only read them explicitly with
>>>> "ctest_read_custom_files(${CTEST_BINARY_DIRECTORY})" in my ctest script.
>>>> My CTestCustom.cmake is generated by an earlier call to
>>>> ctest_configure().
>>>>
>>>> Nils
>>>>
>>>> On 10/03/2012 08:24 PM, David Cole wrote:
>>>>> Yeah, seems silly, I know. But ctest and cmake both need to know about
>>>>> it, and we didn't (and still don't) want to introduce a dependency of
>>>>> reading the CMakeCache file from ctest.
>>>>>
>>>>> Let us know if there's still an issue after you do that.
>>>>>
>>>>>
>>>>> On Wed, Oct 3, 2012 at 1:56 PM, Julien Malik <julien.malik at c-s.fr>
>>>>> wrote:
>>>>>> OK I did not understand I needed *both*.
>>>>>> Reading the blog article again, I feel stupid. I swear I've read it
>>>>>> several
>>>>>> times ;)...
>>>>>>
>>>>>> That must be it.
>>>>>>
>>>>>> Thanks for hint,
>>>>>> Julien
>>>>>>
>>>>>>
>>>>>> Le 03/10/2012 19:25, David Cole a écrit :
>>>>>>
>>>>>>> http://www.kitware.com/blog/home/post/11
>>>>>>>
>>>>>>> Did you try setting CTEST_USE_LAUNCHERS in the ctest -S script, too?
>>>>>>>
>>>>>>> It's in your CMakeCache already, but it is not set when you run the
>>>>>>> outer level ctest script. The driving ctest needs to know to use
>>>>>>> CTEST_USE_LAUNCHERS also because that's what decides whether to
>>>>>>> generate the Build.xml from the logs of the individually launched
>>>>>>> commands, or whether to scrape the log of the mixed up results of all
>>>>>>> the commands from one giant output stream.
>>>>>>>
>>>>>>> That should fix it.
>>>>>>>
>>>>>>>
>>>>>>> HTH,
>>>>>>> David
>>>>>>>
>>>>> --
>>>>>
>>>>> Powered by www.kitware.com
>>>>>
>>>>> Visit other Kitware open-source projects at
>>>>> http://www.kitware.com/opensource/opensource.html
>>>>>
>>>>> Please keep messages on-topic and check the CMake FAQ at:
>>>>> http://www.cmake.org/Wiki/CMake_FAQ
>>>>>
>>>>> Follow this link to subscribe/unsubscribe:
>>>>> http://www.cmake.org/mailman/listinfo/cmake
>>>>
>>>>
>>
>> --
>> Nils Gladitz, B.Sc.
>> DICOM, Konnektivität und Entwicklung
>>
>> Scivis wissenschaftliche Bildverarbeitung GmbH
>> Bertha-von-Suttner-Str. 5
>> D-37085 Göttingen
>> GERMANY
>> Handelsregister Nr. / Trade Register No. B3100 Göttingen
>> Geschäftsführer / Managing Directors Dr. Gernot Ebel, Dr. Uwe Engeland
>>
>> Tel: 0049 (0)551 634181-28
>> E-Mail: gladitz at scivis.de
>> Web: www.scivis.de
>>
>> --
>>
>> Powered by www.kitware.com
>>
>> Visit other Kitware open-source projects at
>> http://www.kitware.com/opensource/opensource.html
>>
>> Please keep messages on-topic and check the CMake FAQ at:
>> http://www.cmake.org/Wiki/CMake_FAQ
>>
>> Follow this link to subscribe/unsubscribe:
>> http://www.cmake.org/mailman/listinfo/cmake


-- 
Nils Gladitz, B.Sc.
DICOM, Konnektivität und Entwicklung

Scivis wissenschaftliche Bildverarbeitung GmbH
Bertha-von-Suttner-Str. 5
D-37085 Göttingen
GERMANY
Handelsregister Nr. / Trade Register No. B3100 Göttingen
Geschäftsführer / Managing Directors Dr. Gernot Ebel, Dr. Uwe Engeland

Tel: 0049 (0)551 634181-28
E-Mail: gladitz at scivis.de
Web: www.scivis.de