LabPlot2 was built with the old system but is not present in the new one yet. Only the master branch should be considered.
Please also active Windows, MacOSX and FreeBSD builds.
Description
Description
Details
Details
- Commits
- R857:2ef933a20cb7: Subscribe Alexander to Labplot build notifications as requested. Fixes T6465
R857:500278d50eec: Add GSL to the Windows nodes for Labplot. Ref T6465
R857:e233f57e3ceb: Labplot needs GSL. Hopefully Labplot doesn't have too many dependencies we are…
R857:5c40e306c245: Enable Windows and FreeBSD builds for Labplot as well. Ref T6465
R857:161cd1dc7a1d: Enable Labplot on the CI system. Ref T6465
Related Objects
Related Objects
- Mentioned In
- R138:f6c669fa023a: Update GSL to 2.2.1
Comment Actions
I've now enabled SUSE (Linux), FreeBSD and Windows builds. Mac OS builds are not available at this time.
However, there are some limitations:
- It takes approximately 30-40 minutes for the Linux image to rebuild. While i've added GSL I don't know if anything else is required because CMake bails on it's first error so other dependencies may be missed. Please consider fixing that. If anything else is missing, patches to add the necessary packages would be appreciated.
- GSL is not currently available on Windows as it fails to compile. Please contact the KDE on Windows / Craft developers (who i've added to this task) to get this resolved. Same issue with CMake dependencies as 1) also applies here. The GSL failure log is as follows:
[ 80%] Building C object CMakeFiles/sys_test.dir/sys/test.c.obj test.c [ 80%] Linking C executable sys_test.exe gsl.lib(pow_int.c.obj) : error LNK2005: gsl_pow_2 already defined in test.c.obj gsl.lib(pow_int.c.obj) : error LNK2005: gsl_pow_3 already defined in test.c.obj gsl.lib(pow_int.c.obj) : error LNK2005: gsl_pow_4 already defined in test.c.obj gsl.lib(pow_int.c.obj) : error LNK2005: gsl_pow_5 already defined in test.c.obj gsl.lib(pow_int.c.obj) : error LNK2005: gsl_pow_6 already defined in test.c.obj gsl.lib(pow_int.c.obj) : error LNK2005: gsl_pow_7 already defined in test.c.obj gsl.lib(pow_int.c.obj) : error LNK2005: gsl_pow_8 already defined in test.c.obj gsl.lib(pow_int.c.obj) : error LNK2005: gsl_pow_9 already defined in test.c.obj sys_test.exe : fatal error LNK1169: one or more multiply defined symbols found LINK Pass 1 failed. with 1169 NMAKE : fatal error U1077: 'C:\Craft\dev-utils\cmake\bin\cmake.exe' : return code '0xffffffff' Stop. NMAKE : fatal error U1077: '"C:\Program Files (x86)\Microsoft Visual Studio\2017\Enterprise\VC\Tools\MSVC\14.10.25017\bin\HostX64\x64\nmake.exe"' : return code '0x2' Stop. NMAKE : fatal error U1077: '"C:\Program Files (x86)\Microsoft Visual Studio\2017\Enterprise\VC\Tools\MSVC\14.10.25017\bin\HostX64\x64\nmake.exe"' : return code '0x2' Stop. Craft encountered an error: make cmd: nmake /NOLOGO Action: compile for win32libs/gsl FAILED *** Craft all failed: all of win32libs/gsl failed after 0:02:59 *** fatal error: package win32libs/gsl all failed
- For FreeBSD, no Extragear project to date has requested a build there, so the system needs to compile the base (Frameworks and all other non-Extragear libraries which are required by Extragear applications who are enabled for FreeBSD builds). Once this has been completed we can try a first build of Labplot on FreeBSD.
Comment Actions
- Linux build is working fine. I fixed CMake to not bail out on first error.
- GSL is now available on Windows and the build starts. We never used MSVC before, so we need to fix a few things in the code :-)
- FreeBSD build now fails with missing GSL. Can you add GSL also to FreeBSD?
Comment Actions
- Please add @asemke to the list of people receiving emails from CI.
- The Windows build is not active anymore. Is there any problem?
Comment Actions
I'll do 1) shortly.
In terms of 2) it is very much still active, although it has never succeeded (MSVC issues it would seem): https://build.kde.org/view/Extragear/job/Extragear%20labplot%20kf5-qt5%20WindowsMSVCQt5.9/19/console
Comment Actions
thanks for 1)
For 2) i was wondering why the latest build is from August. It seems that git commits are not triggering a new build.
Comment Actions
Did you change branch by any chance?
Changes to branches mean the system won't pick it up until a rebuild has been done manually.
I've now kicked off rebuilds for all the Labplot builds.