Re: [CMake] Controlling output of ctest

2014-08-26 Thread Ashok Nalkund
I think I've found the solution at http://www.vtk.org/Wiki/CMake_Testing_With_CTest. I'm going to try setting CTEST_CUSTOM_MAXIMUM_PASSED_TEST_OUTPUT_SIZE and CTEST_CUSTOM_MAXIMUM_FAILED_TEST_OUTPUT_SIZE and report back. Thanks, Ashok On 8/26/2014 6:24 PM, Ashok Nalkund wrote: H

[CMake] Controlling output of ctest

2014-08-26 Thread Ashok Nalkund
Hi, I use cmake/ctest to build and run my tests. I'm using "ctest -T Test -V --output-on-failure --no-compress-output -R ...". I noticed that some of the tests have truncated output in the Test.xml report. Also I see the truncation length is different in different scenarios. From what I see,

[CMake] cmake's REMOVE_RECURSE is not deleting symlinks

2014-05-20 Thread Ashok Nalkund
Hi All, I have a situation where I need to: - create some dirs in the build directory - create symlinks to the above dirs in the source directory The creation part works well, I'm using add_custom_command(OUTPUT ...) and add_custom_target() which depends onthe output of custom command. The

[CMake] Force single threaded build in a directory

2013-09-04 Thread Ashok Nalkund
Hi All, I have a package which gives strange results when I build it with multiple threads (make -j...). Is there a way I can force the package to be built using a single thread? Something I can put in the CMakeLists.txt? thanks, ashok -- Powered by www.kitware.com Please keep messages on

Re: [CMake] ExternalProject feature regenerating the configure script

2013-07-30 Thread Ashok Nalkund
On 7/30/2013 12:37 PM, Ashok Nalkund wrote: On 7/30/2013 8:36 AM, David Cole wrote: > The never-ending debate about fixes versus workarounds starts up > again. :-) > ... > ... for everybody in the future who > would otherwise be tripping over this issue. So I hope he goe

Re: [CMake] ExternalProject feature regenerating the configure script

2013-07-30 Thread Ashok Nalkund
On 7/30/2013 8:36 AM, David Cole wrote: > The never-ending debate about fixes versus workarounds starts up > again. :-) > ... > ... for everybody in the future who > would otherwise be tripping over this issue. So I hope he goes for > the fix rather than the workaround, but that is obviously u