+++ This bug was initially created as a clone of Bug #19781 +++ - Architecture [also part of the ISO name below]: - i586 Intel chip set 3.0 Ghz Intel Pentium 4 HT processor 2 Gigs of Ram on 400 MHz in dual channel - Intel Graphics with memory bus sharing i865G both drivers - what is available on download link - BIOS boot - At what stage do you see the message? - Booting the ISO from [what medium] - Later during the installation - Re-booting the installed system: - Without non-free [video] software - With non-free [video] software - Without installation update - Kernel version(s) all - What graphics driver? both non free and free - Which display manager? Default on install I have tried to use drakconf with no success. I got error EE from drakconf. If this is happening with more than one video card I would double check that the hardware support is compiled in kernel. Their is two option when last time I compiled the kernel in UNIX. You have a choice of built in the kernel and dynamic. The dynamic should be used. This option loads hardware support when detected. From the looks of it I see in mga5 when I changed video cards it had hardware support already and installed drivers. I would look at the kernel first in my opinion. I would also check bios setting too if applicable but is not in my case. the display actually starts but is black with no back light effect. move the mouse to the left. you will see the mouse arrow. wait some time and it turns into a x.
Please switch to a VT with ctrl+alt+F3, login as root and run journalctl -ab > journal.txt and attach journal.txt to this bug report
Keywords: TRACKER => (none)
(when you hit this bug, of course ;-) )
URL: https://bugs.mageia.org/show_bug.cgi?id=19216 => (none)See Also: (none) => https://bugs.mageia.org/show_bug.cgi?id=19216
re-installing version text from text file Mageia Mageia vanda 20160629 19:33 i hava had similar effect on mga5 when enabling 3d graphics compiz fusion. I seen a Glibc error with qt3 on mga5 will see what happens when I install cauldon again.
new keyboard sorry, cauldron 6 :>
(In reply to Marja van Waes from comment #1) > > Please switch to a VT with ctrl+alt+F3, login as root and run > > journalctl -ab > journal.txt > > and attach journal.txt to this bug report Once I have done that how do I attach it to bug report? No display and no way to get it to attach to bug report. lynx VT browser hmmm. new user will not know this or how to use lynx. Your in luck as I have re-sized mga5 partition and installed cauldron 6 sta1 on second partition. I had to mount the cauldron partition with root permission new one. Now I tried to open location with log and got error message. I displayed the VT terminal on the view panel option. I su to root cp logfile to my home dir. Then chown the file to me as user in order for me to get this file attached here. Your lucky I have some good knowledge of UNIX and Linux. :~) I hope this helps. have a good morning :>
Created attachment 8674 [details] Journal log requested.
@ Stephen Thanks for having figured out how to supply that log file :-) Please give the following information, too: * Tell again (I didn't understand) how exactly you installed Mageia 6. * if you used a traditional, non-Live iso: ** attach /root/drakx/report.bug.xz from that install * however, if you used a Live iso: ** attach install.txt that is the result of (as root): journalclt -ab1 > install.txt ** and run (it's OK to do that in a working Mageia on the same hardware) lspcidrake -v > lspci.txt and attach lspci.txt, too About attachment 8674 [details] : Nov 16 22:56:09 StephenPCL sddm-greeter[1836]: Message received from daemon: Capabilities Nov 16 22:56:09 StephenPCL sddm-greeter[1836]: Message received from daemon: HostName Nov 16 22:56:09 StephenPCL sddm-greeter[1836]: QOpenGLShader: could not create shader Nov 16 22:56:09 StephenPCL sddm-greeter[1836]: QOpenGLShader: could not create shader Nov 16 22:56:32 StephenPCL audit[1836]: ANOM_ABEND auid=4294967295 uid=994 gid=990 ses=4294967295 pid=1836 comm="sddm-greeter" exe="/usr/bin/sddm-greeter" sig=11 Nov 16 22:56:32 StephenPCL kernel: show_signal_msg: 102 callbacks suppressed Nov 16 22:56:32 StephenPCL kernel: sddm-greeter[1836]: segfault at 0 ip (null) sp bff0627c error 4 in sddm-greeter[8048000+46000] Nov 16 22:56:32 StephenPCL kernel: audit: type=1701 audit(1479358592.782:121): auid=4294967295 uid=994 gid=990 ses=4294967295 pid=1836 comm="sddm-greeter" exe="/usr/bin/sddm-greeter" sig=11 Nov 16 22:56:32 StephenPCL sddm-helper[1756]: [PAM] Ended. Nov 16 22:56:32 StephenPCL sddm[1527]: Auth: sddm-helper exited with 11 Nov 16 22:56:32 StephenPCL sddm[1527]: Greeter stopped. Assigning to SDDM This looks like a duplicate of bug 16544, but let's wait for Stephen's feedback before deciding it's the same bug or not. @ Stephen Can you please also report whether adding Option "DRI" "false" in the "Device" section for your graphic card solves the issue?
Keywords: (none) => NEEDINFOSee Also: (none) => https://bugs.mageia.org/show_bug.cgi?id=16544, https://github.com/sddm/sddm/issues/376Assignee: bugsquad => kdeSummary: Mageia 6 REAL HARDWARE: "Good luck" failure to start graphical interface, display manager => sddm-greeter segfaults after "sddm-greeter: QOpenGLShader: could not create shader" (another "Good luck" failure)Source RPM: (none) => sddm
(In reply to Marja van Waes from comment #7) > > @ Stephen > > Can you please also report whether adding > > Option "DRI" "false" > > in the "Device" section for your graphic card solves the issue? That's in /etc/X11/xorg.conf
(In reply to Marja van Waes from comment #8) > (In reply to Marja van Waes from comment #7) > > > > > @ Stephen > > > > Can you please also report whether adding > > > > Option "DRI" "false" > > > > in the "Device" section for your graphic card solves the issue? > > That's in /etc/X11/xorg.conf I installed with DVD-iso 32 bit and that did get me to a login screen Have a nice day ;)
I have logged into mga6 ok looks ok. Mageia 6 kernel: 4.6.3-desktop-1.mga6 arch: 32-bit Desktop: KDE looking into kernel as soon as I create the developers environment. haven't seen the various drivers for video graphics from kernel. This means that it must be built in to the kernel as I only see generic video kernel drivers.
I have fully updated my system and the newest intel 810 and later driver install rewrite of config file same results. even with the newer kernel. Both original and newer kernels detects the wrong intel video card. I get memory error for a i915 video card but mine is i865. When i get home scratch that downloading src rpm now because src rpm for original cauldron kernel was not on mirror. I will extract the kernel source code from kernel and look at the driver and video section for intel for coding errors. I hope all the video cards are not built into the kernel instead should be built as insertable mod. The reason for this is to prevent issues like this or catch them faster. Mageia 6 kernel: 4.8.8-desktop-1.mga6 arch: 32-bit Desktop: KDE
@ Stephen Thanks for all the feedback.... can you please elaborate on: (In reply to Stephen Butler from comment #9) > (In reply to Marja van Waes from comment #8) > > (In reply to Marja van Waes from comment #7) > > > > > > > > @ Stephen > > > > > > Can you please also report whether adding > > > > > > Option "DRI" "false" > > > > > > in the "Device" section for your graphic card solves the issue? > > > > That's in /etc/X11/xorg.conf > > I installed with DVD-iso 32 bit Which DVD-iso 32 bit * Mageia 6 sta 1 ? * classical (so non-Live)? ** if Live, then Plasma5? > > and that did get me to a login screen To avoid misunderstanding: What is "that" in the above sentence? Adding 'Option "DRI" "false"' in /etc/X11/xorg.conf ? Also, please attach the one or two files (depending on used installation method) requested in comment 7: > > * if you used a traditional, non-Live iso: > ** attach /root/drakx/report.bug.xz from that install > > * however, if you used a Live iso: > ** attach install.txt that is the result of (as root): > journalclt -ab1 > install.txt > ** and run (it's OK to do that in a working Mageia on the same hardware) > lspcidrake -v > lspci.txt > and attach lspci.txt, too > Thanks for your help :-)
Priority: release_blocker => Normal
> > Option "DRI" "false" and that did get me to a login screen Have a nice day ;) Which DVD-iso 32 bit * Mageia 6 sta 1 ? * classical (so non-Live)? ** if Live, then Plasma5? you have three choices on download screen. confusing but I needed Full DVD so the classical. 32 bit. BitTorrent. upload install logs tonight when I get home.
(In reply to Stephen Butler from comment #13) > > upload install logs tonight when I get home. Good :-)
Created attachment 8677 [details] install logs
Created attachment 8678 [details] hardware logs
Thanks for all the feedback, Stephen :-) I hope to do a fresh install on the system where I hit bug 16544, soon, to check whether I get the "Good luck" error too, now, instead of a cursor on a black screen.
Keywords: NEEDINFO => (none)
I see their is allot of kernel patches in the source file and incomplete kernel sources from source rpm. I will compile the kernel from scratch and see if that fixes anything.
(In reply to Stephen Butler from comment #18) > I see their is allot of kernel patches in the source file and incomplete > kernel sources from source rpm. I will compile the kernel from scratch and > see if that fixes anything. This bug is getting into deep water! Some basic points: = The starting point was M6 sta1. So old as to no longer be relevant. Noted that you fully updated it. We have suspended M6 ISO testing until new ISOs are available after releasing 5.1. It might be better to await them, and re-try then. Unless the problem is sorted sooner. = SDDM (intimately associated with Plasma) caused a lot of problems earlier on, but there are some pre-emptive M6 testers who now have good systems using it. Did you ever try *changing the display manager*, which can be done from console using # drakdm ? LXDM and LightDM are the surest. This seems vital. Another ruse while installing from the Classic ISO is to choose Mate, LXDE or XFCE desktops, not Plasma/KDE, which will automatically install & default to a different Display Manager. = If we understand correctly from your Comment 13 that: in /etc/X11/xorg.conf in the "Device" section for your graphic card adding 'Option "DRI" "false"' DID resolve the problem, that is important to know. This would be another known possible remedy for the "Good luck" problem. And effectively resolve your problem. Or is there something more? = Your initial report said the fault arose with "both non free and free" video drivers. So it might be easier to stick with 'free' to persue the problem, since using nonfree adds complications. OTOH If the problem only happens with free, and *not* with nonfree, it shifts the focus. = Your Comment 11 suggests that this bug is graphic hardware sensitive. The issue of the wrong video h/w being reported could be a separate bug - if it matters. Would not both the i865 and i915 be accomodated by "the newest intel 810 and later driver"? Do you have the means to try different video hardware? For the moment, it looks rather ambitious to be compiling kernels to investigate this. But if you find something, so much the better.
I did try to install the Gnome display manager. I got login screen but tried to login to each of the three didn't work. I tried using ATI radeon 128 card and it didn't work very well.
I have configured the kernel and enabled some stuff that is for graphics card support. The DRM_I915 supports i810 and later, But their is also I810 and later build for DRM_I810. ************************* Configuration for kernel 4.8.8 build for by default DRM_810=y DRM_915=y ************************* I have chosen not to compile DRM_810 as DRM_915 supports 810 and later graphic cards. If it works then will test options "DRI" "true".
Stephen, is the hardware you're using is 32bit, or is it 64bit capable? (Asking because the hardware used in probable duplicate bug 16544 cannot handle 64bit, and because I'm not aware anyone reported seeing "sddm-greeter: QOpenGLShader: could not create shader" in a 64bit install.... if your hw can handle 64bit, it would be interesting to see whether the problem occurs in a 64bit install, too)
There has been a fix for most "good luck" issues, is this bug report still valid or is it fixed now?
Keywords: (none) => NEEDINFO
No reply, but given the same error message that precedes the segfault, and the fact that the workaround from bug 16544 solved the problem, it must be a duplicate *** This bug has been marked as a duplicate of bug 16544 ***
Status: NEW => RESOLVEDResolution: (none) => DUPLICATE