KDE Gear projects with failing CI (master) (18 June 2024)

Ingo Klöcker kloecker at kde.org
Fri Jun 21 08:29:18 BST 2024


On Freitag, 21. Juni 2024 00:36:45 CEST Albert Astals Cid wrote:
> El dimecres, 19 de juny del 2024, a les 9:07:50 (CEST), Ingo Klöcker va
> 
> escriure:
> > On Mittwoch, 19. Juni 2024 00:49:14 CEST Albert Astals Cid wrote:
> > > kdenlive - 2nd week
> > > 
> > >  * https://invent.kde.org/multimedia/kdenlive/-/pipelines/716283
> > >  
> > >   * craft_appimage_qt6_x86_64 fails
> > > 
> > > isoimagewriter - NEW
> > > 
> > >  * https://invent.kde.org/utilities/isoimagewriter/-/pipelines/716655
> > >  
> > >   * craft_appimage_qt6_x86_64 fails
> > > 
> > > neochat - NEW
> > > 
> > >  * https://invent.kde.org/network/neochat/-/pipelines/716759
> > >  
> > >   * craft_appimage_qt6_x86_64 fails
> > 
> > All of these fail because the tests of kconfig don't build. I'm wondering
> > why not all AppImage builds of all projects fail.
> > 
> > The cause for the failing (or highly unstable) builds of the tests of
> > kconfig is an error in the test design. Here's what I wrote late last
> > night
> > in #kde-craft:kde.org.
> > 
> > ===
> > There is a pair of tests, e.g. Test8 and Test8_cmake where the latter is
> > supposed to test that cmake can generate a .kcfgc file. The problem is
> > that
> > both tests share the same test8main.cpp which includes a test8a.h. This
> > test8a.h is generated via test8a.kcfg[c] for Test8, but for Test8_cmake
> > test8a_cmake.h is generated via a test8a_cmake.kcfg[c]. If for some reason
> > Test8_cmake is built before Test8 then test8main.cpp doesn't find
> > test8a.h.
> > I'm not yet sure how to fix this. Maybe we need to create patched copies
> > of
> > testXmain.cpp which includes a testXx_cmake.h instead of a testXx.h. Or
> > maybe we put both tests in separate subdirs and don't append _cmake to the
> > files used for the TestX_cmake tests. That's feels like a cleaner solution
> > also because it better isolates the two tests from each other.
> > ===
> > 
> > With make the problem doesn't seem to occur because it looks like the
> > Makefiles serialize building of Test8 and Test8_cmake so that Test8_cmake
> > is always built after Test8 where Test8 creates the headers that
> > Test8_cmake uses. With ninja the builds of both targets don't seem to be
> > serialized so that Test8_cmake could be built before Test8. -> Fail!
> 
> With that info, we got a fix in, but that won't fix CI since it uses
> tarballs, should we get
> https://invent.kde.org/frameworks/kconfig/-/merge_requests/314/diffs
> added as a patch to Craft?

In the meantime building of the KConfig tests has been disabled in Craft. We 
can just wait for KF 6.4.

Actually, I'm wondering why we build tests in Craft at all. Craft doesn't run 
the tests, the tests are not packaged _and_ the purpose of the Craft jobs 
isn't CI. I'd say we should probably disable building of tests in general in 
Craft. Building them is just a waste of energy and time and I fail to see any 
benefit. When/if we want to start running the tests in Craft then we can simply 
re-enable their build.

Regards,
Ingo
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 228 bytes
Desc: This is a digitally signed message part.
URL: <http://mail.kde.org/pipermail/kde-devel/attachments/20240621/8ce7196c/attachment.sig>


More information about the kde-devel mailing list