[cmake-developers] [CMake 0014414]: Ninja: Running compiler from build dir makes it being run from bad relative directory
The following issue has been SUBMITTED. == http://www.cmake.org/Bug/view.php?id=14414 == Reported By:Damian Kaczmarek Assigned To: == Project:CMake Issue ID: 14414 Category: CMake Reproducibility:always Severity: block Priority: normal Status: new == Date Submitted: 2013-09-16 03:15 EDT Last Modified: 2013-09-16 03:15 EDT == Summary:Ninja: Running compiler from build dir makes it being run from bad relative directory Description: During my build setup I am creating build/bin/ccache-arm-none-eabi-gcc My toolchain setup sets this binary: set(CMAKE_C_COMPILER ${CCACHE_GCC_BINARY} CACHE FORCE ) Everything is fine when using make, however when using ninja the following problem arose: [6/1347] Building ASM object Project/CMakeFiles/Project_startup.dir/startup.s.obj FAILED: ../bin/ccache-arm-none-eabi-gcc ... rest of command ... /bin/sh: ../bin/ccache-arm-none-eabi-gcc: No such file or directory I may try to provide a small test project setup if requested. == Issue History Date ModifiedUsername FieldChange == 2013-09-16 03:15 Damian KaczmarekNew Issue == -- 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://public.kitware.com/cgi-bin/mailman/listinfo/cmake-developers
[cmake-developers] AUTOUIC and AUTORCC for Qt
Hi there, I've pushed the Qt-auto-generators topic to my clone. https://gitorious.org/cmake/steveires-cmake/commits/Qt-auto-generators It is a proof of concept for features similar to AUTOMOC, but for invoking the uic and rcc tools shipped with Qt. I'd like to generalize the AUTOMOC feature to be able to remove more of the need to use macros with Qt for ui files and rcc files. Any comments? Thanks, Steve. -- 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://public.kitware.com/cgi-bin/mailman/listinfo/cmake-developers
[cmake-developers] [CMake 0014415]: FindCUDA.cmake INCLUDE_DIRECTORIES add/honor optional SYSTEM flag
The following issue has been SUBMITTED. == http://www.cmake.org/Bug/view.php?id=14415 == Reported By:Axl Assigned To: == Project:CMake Issue ID: 14415 Category: Modules Reproducibility:always Severity: feature Priority: normal Status: new == Date Submitted: 2013-09-16 08:40 EDT Last Modified: 2013-09-16 08:40 EDT == Summary:FindCUDA.cmake INCLUDE_DIRECTORIES add/honor optional SYSTEM flag Description: Like GCC, nvcc knows and honors the -isystem include flag instead of -I . It would be nice, if the CUDA_NVCC_INCLUDE_ARGS would not copy all includes to -I but use the appropriate isystem flag instead if it was set before: http://cmake.org/gitweb?p=cmake.git;a=blob;f=Modules/FindCUDA.cmake;h=0390ae4ad78447c0ce1cb32a899e9d0f1218d952;hb=HEAD#l1040 Steps to Reproduce: Write an example CUDA program using, e.g. boost program options. ... FIND_PACKAGE(Boost REQUIRED COMPONENTS program_options) INCLUDE_DIRECTORIES(SYSTEM ${Boost_INCLUDE_DIRS}) ... make VERBOSE=1 FindCUDA will add all SYSTEM libraries with -I instead of -isystem when calling nvcc. Additional Information: Thats the line when it happens: http://cmake.org/gitweb?p=cmake.git;a=blob;f=Modules/FindCUDA.cmake;h=0390ae4ad78447c0ce1cb32a899e9d0f1218d952;hb=HEAD#l1040 == Issue History Date ModifiedUsername FieldChange == 2013-09-16 08:40 AxlNew Issue == -- 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://public.kitware.com/cgi-bin/mailman/listinfo/cmake-developers
Re: [cmake-developers] Improving GraphViz output, clustering
On Mon, Sep 16, 2013 at 7:39 PM, Aleix Pol aleix...@kde.org wrote: Hi, I've been trying to get a good outlook of the internal dependencies of one of our projects (KDE Frameworks 5). One thing that I've found useful is adding clustering to external dependencies that, while needed in the graph, having them grouped as a separate entity is desirable. Attached, I'm providing a patch that adds this functionality. I hope you find it useful. Aleix And of course, I forgot the attachment. :) Aleix -- 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://public.kitware.com/cgi-bin/mailman/listinfo/cmake-developers
[cmake-developers] Improving GraphViz output, clustering
Hi, I've been trying to get a good outlook of the internal dependencies of one of our projects (KDE Frameworks 5). One thing that I've found useful is adding clustering to external dependencies that, while needed in the graph, having them grouped as a separate entity is desirable. Attached, I'm providing a patch that adds this functionality. I hope you find it useful. Aleix -- 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://public.kitware.com/cgi-bin/mailman/listinfo/cmake-developers
[cmake-developers] [CMake 0014417]: cmake --system-information sfile .. removes x perm from ../ and fails
The following issue has been SUBMITTED. == http://public.kitware.com/Bug/view.php?id=14417 == Reported By:Steven R Thompson Assigned To: == Project:CMake Issue ID: 14417 Category: CMake Reproducibility:always Severity: minor Priority: normal Status: new == Date Submitted: 2013-09-16 14:39 EDT Last Modified: 2013-09-16 14:39 EDT == Summary:cmake --system-information sfile .. removes x perm from ../ and fails Description: from my build subdir, I executed cmake --system-information sinfo .. and it generated lots of error messages like this: CMake Error st CMakeLists.txt:76 (FILE): file Internal CMake error when trying to open file: /usr/people/thompson/Xdfm2/build/.. for writing Call Stck (most recent call first): CMakeLists.txt(83) (DUMP_FILE) the final error message is: CMake Warning (dev) in CMakeLists.txt: No cmake_minimum_required command is present. A line of code such as cmake_minimum_required(VERSION 2.6) should be added to the top of the file. After running this, I was not able to execute ls .., due to the fact that x-permission had been removed from the parent directory. I was able to reproduce this several times, after manually restoring x-permission each time. Apparently it removes x-permission early and then cannot access the files in the parent directory, which results in all the error messages. Steps to Reproduce: cmake --system-information sinfo .. Additional Information: Also, despite the error about there being no cmake_minimum_required line, it is in the file ../CMakeLists.txt. I ran the command with .. as the final argument since the man page for cmake indicates to me that it is required (and other cmake commands require a directory as an arg). Since the x-permission as removed from .., I tried running cmake without .. as an arg, and got different results. The x-permission was not removed, but I still got the warning message about cmake_minimum_required. I did get an sinfo file, so the command worked better without .. as an arg. So at this point I would say that it gives an erroneous warning, that the man page is misleading if a subdirectory is given, and that if a subdirectory is given, it should not remove x-permission. == Issue History Date ModifiedUsername FieldChange == 2013-09-16 14:39 Steven R ThompsonNew Issue == -- 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://public.kitware.com/cgi-bin/mailman/listinfo/cmake-developers
Re: [cmake-developers] Improving GraphViz output, clustering
On Monday 16 September 2013, Aleix Pol wrote: On Mon, Sep 16, 2013 at 7:39 PM, Aleix Pol aleix...@kde.org wrote: Hi, I've been trying to get a good outlook of the internal dependencies of one of our projects (KDE Frameworks 5). One thing that I've found useful is adding clustering to external dependencies that, while needed in the graph, having them grouped as a separate entity is desirable. Attached, I'm providing a patch that adds this functionality. I hope you find it useful. Aleix And of course, I forgot the attachment. :) Hmm, I still don't see the attachement... Alex -- 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://public.kitware.com/cgi-bin/mailman/listinfo/cmake-developers
Re: [cmake-developers] Improving GraphViz output, clustering
On Mon, Sep 16, 2013 at 10:03 PM, Alexander Neundorf neund...@kde.orgwrote: On Monday 16 September 2013, Aleix Pol wrote: On Mon, Sep 16, 2013 at 7:39 PM, Aleix Pol aleix...@kde.org wrote: Hi, I've been trying to get a good outlook of the internal dependencies of one of our projects (KDE Frameworks 5). One thing that I've found useful is adding clustering to external dependencies that, while needed in the graph, having them grouped as a separate entity is desirable. Attached, I'm providing a patch that adds this functionality. I hope you find it useful. Aleix And of course, I forgot the attachment. :) Hmm, I still don't see the attachement... Alex I uploaded the patch, gmail seems to hate me today... My apologies. http://proli.net/meu/kde/cmake-clusteredgraphs.patch Aleix -- 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://public.kitware.com/cgi-bin/mailman/listinfo/cmake-developers
[cmake-developers] [CMake 0014418]: FindOpenSSL: should mark {LIB, SSL}_EAY_{DEBUG, RELEASE} as advanced
The following issue has been SUBMITTED. == http://public.kitware.com/Bug/view.php?id=14418 == Reported By:Daniel Schepler Assigned To: == Project:CMake Issue ID: 14418 Category: CMake Reproducibility:always Severity: minor Priority: normal Status: new == Date Submitted: 2013-09-16 20:12 EDT Last Modified: 2013-09-16 20:12 EDT == Summary:FindOpenSSL: should mark {LIB,SSL}_EAY_{DEBUG,RELEASE} as advanced Description: After running find_package(OpenSSL) on Windows, the variables LIB_EAY_DEBUG, LIB_EAY_RELEASE, SSL_EAY_DEBUG, SSL_EAY_RELEASE show up in the CMake GUI as non-advanced variables. (In my setup, it was not found -- but looking at FindOpenSSL.cmake, I don't see any mark_as_advanced() with these variables in any case.) Steps to Reproduce: find_package(OpenSSL) == Issue History Date ModifiedUsername FieldChange == 2013-09-16 20:12 Daniel ScheplerNew Issue == -- 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://public.kitware.com/cgi-bin/mailman/listinfo/cmake-developers