Am Donnerstag, 30. Juli 2015, 21:04:45 schrieb Dominik George: > Hi, > > > I suggest taking this upstream to https://bugs.kde.org, cause it really > > sounds like an upstream bug. > > I will look into this. Also please note that Qt 5.5 and even more so 5.5.1 will likely will likely fix a lot of multi screen related issues. It is not yet in unstable and not yet complete. There is a gcc transition taking place and Lisandro intends to get it in after this transition. > > So please also > > install the necessary -dbg packages and try to create a backtrace. > > What I can provide now is this in dmesg: > > [14085.109351] [drm:intel_set_cpu_fifo_underrun_reporting [i915]] *ERROR* > uncleared fifo underrun on pipe A > [14085.109370] [drm:ironlake_irq_handler [i915]] *ERROR* CPU pipe A FIFO > underrun > [14090.111493] kactivitymanage[24108]: segfault at 7fb50847ecd0 ip > 00007fb508426031 sp 00007ffc5e1d3b48 error 4 in > libQt5Sql.so.5.4.2[7fb508412000+3f000] > [ 86.841807] kscreen_backend[2331]: segfault at 10 ip 00007f5fea953bd0 sp > 00007ffd093cb508 error 4 in KSC_XRandR.so[7f5fea93e000+22000] > [ 87.698749] ksplashqml[2280]: segfault at 8 ip 00007fe9533685a0 sp > 00007ffc6aca1e38 error 4 in libQt5Gui.so.5.4.2[7fe953256000+52c000] > [ 87.701634] kactivitymanage[2324]: segfault at 18 ip 00007f4538f7e910 sp > 00007ffdf990aec8 error 4 in libqxcb.so[7f4538f4f000+ba000] > [ 87.728423] kscreen_backend[2340]: segfault at 18 ip 00007f5df5a80b29 sp > 00007fff090273a0 error 4 in libqxcb.so[7f5df5a36000+ba000] > [ 87.860003] kaccess[2343]: segfault at 18 ip 00007fb2b9cb2b29 sp > 00007ffced3b94d0 error 4 in libqxcb.so[7fb2b9c68000+ba000] > [ 88.159947] ksmserver[2352]: segfault at 18 ip 00007f1551262b29 sp > 00007ffece1b7a20 error 4 in libqxcb.so[7f1551218000+ba000] > [ 92.911895] [drm:intel_set_cpu_fifo_underrun_reporting [i915]] *ERROR* > uncleared fifo underrun on pipe A > [ 92.911963] [drm:ironlake_irq_handler [i915]] *ERROR* CPU pipe A FIFO > underrun > > > This looks like something seriously drowns ;). Also, it remotely suggests > that the Intel DRM component might be involved. Hmm, yes. I have no idea about it. Would be good for an upstream bug report I think. > > You may even report upstream by using DrKonqi. > > Can this report an issue that happened in the past? Hmmm, I used DrKonqi only as it was popped automatically. So you don´t get a DrKonqi at the crash at startup? It would help creating backtraces. Anyway, it may be wise to just retest with Qt 5.5. Maybe the issue is fixed already. Thanks, -- Martin
Attachment:
signature.asc
Description: This is a digitally signed message part.