Description of problem: I updated from Mageia 5 to Mageia 6 using mgaapplet. There were two messages about gvfs: i586 version clashes with installed x86_64 version. I clicked continue. After a reboot: the cauldron appears with messages about building nvidia-current and nvidia304, followed by a black screen. After another reboot: cauldron is flooed by black screen. Pressing the power button causes a power down message and system turns off (not a hard power down). Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: 1. Click on mgaapplet and click "Next" to update system to Mageia 6 2. Reboot and get a black screen: no cursor My hardware: 3.5 GHz Intel 4th Gen Core i7 4770K 84W HD4600 8MB Quad Core CPU Gigabyte GA-Z87X-UD4H LGA1150 ATX Motherboard Gigabyte NVIDIA GTX770 OC 2GB Windforce Version, GV-N77O0C-2GD Messages in /var/log/messages: nvidia: loading out-of-tree module taints kernel. nvidia: module license 'NVIDIA' taints kernel. nvidia-nvlink: Nvlink Core is being initialized, major device number 243 nvidia 0000:01:00.0: vgaarb: changed VGA decodes: olddecodes=io+mem,decodes=none:owns=io+mem NVRM: loading NVIDIA UNIX x86_64 Kernel Module 390.87 Tue Aug 21 12:33:05 PDT 2018 (using threaded interrupts) ... resource sanity check: requesting [mem 0x000e0000-0x000fffff], which spans more than PCI Bus 0000:00 [mem 0x000e0000-0x000e3fff window] caller _nv029980rm+0x57/0x90 [nvidia] mapping multiple BARs resource sanity check: requesting [mem 0x000c0000-0x000fffff], which spans more than PCI Bus 0000:00 [mem 0x000d0000-0x000d3fff window] caller _nv001169rm+0xe3/0x1d0 [nvidia] mapping multiple BARs nvidia-modeset: Loading NVIDIA Kernel Mode Setting Driver for UNIX platforms 390.87 Tue Aug 21 16:16:14 PDT 2018 The "Device" and "Screen" sections in xorg.conf are unchanged from Mageia 5: Section "Device" Identifier "device1" VendorName "NVIDIA Corporation" BoardName "NVIDIA GeForce 400 series and later" Driver "nvidia" Option "DPMS" Option "AddARGBGLXVisuals" Option "DynamicTwinView" "false" EndSection Section "Screen" Identifier "screen1" Device "device1" Monitor "monitor1" DefaultColorDepth 24 Option "metamodes" "nvidia-auto-select +0+0 { ForceCompositionPipeline=On }" EndSection Motherboard graphics are disabled in the motherboard settings.
CC: (none) => martin
Hi Martin, Thanks for the report. When you get the black screen, can you then press "Ctrl Alt F2" simultaneously to get to a txt login prompt? If so, then please become root and run: journalctl -ab > boot.txt and journalctl -a --since="YYYY-MM-DD hh:mm" --until="YYYY-MM-DD hh:mm" > log.txt after adjusting the --since time to right before you started to upgrade and the --until time to right after you finished upgrading. Then please attach both log.txt and boot.txt to this bug report. (In reply to Martin Ward from comment #0) > Description of problem: > > I updated from Mageia 5 to Mageia 6 using mgaapplet. > There were two messages about gvfs: i586 version clashes with > installed x86_64 version. I clicked continue. > I fail to see why it tried to install the 32bit gvfs version on a 64bit system What is the output now of rpm -qa --last | grep x86_64 | wc -l and of rpm -qa --last | grep i586 | wc -l
Keywords: (none) => NEEDINFOCC: sysadmin-bugs => mageiatools, marja11, pkg-bugsComponent: Release (media or process) => RPM Packages
Thanks for your response. Note that I am installing on a separate partition (on a separate disk) which is a clone of the Mageia 5 partition. I am therefore still running on a copy of the Mageia 5 installation. So if you need to know anything about the system before the upgrade, I can help! I tried switching from nvidia to nouveau in xorg.conf and blacklisting nvidia. Now the bubbling cauldron comes up at the correct resolution, but I still get a black screen. The black screen has a text cursor in the top left corner. "Ctrl Alt F2" does get to a login prompt and the screen is at the right resolution (i.e small and not fuzzy text). rpm -qa --last | grep x86_64 | wc -l returns: 3148 rpm -qa --last | grep i586 | wc -l returns: 284 Note that "rpm -qa --last | grep gvfs" on Mageia 5 (before upgrade) gives: gvfs-smb-1.22.3-2.2.mga5.x86_64 Wed 03 Jan 2018 11:25:20 GMT gvfs-mtp-1.22.3-2.2.mga5.x86_64 Wed 03 Jan 2018 11:25:20 GMT gvfs-iphone-1.22.3-2.2.mga5.x86_64 Wed 03 Jan 2018 11:25:20 GMT gvfs-gphoto2-1.22.3-2.2.mga5.x86_64 Wed 03 Jan 2018 11:25:20 GMT gvfs-goa-1.22.3-2.2.mga5.x86_64 Wed 03 Jan 2018 11:25:20 GMT gvfs-fuse-1.22.3-2.2.mga5.x86_64 Wed 03 Jan 2018 11:25:20 GMT gvfs-archive-1.22.3-2.2.mga5.x86_64 Wed 03 Jan 2018 11:25:20 GMT gvfs-1.22.3-2.2.mga5.x86_64 Wed 03 Jan 2018 11:25:20 GMT libgvfscommon0-1.18.3-1.mga4.i586 Mon 03 Feb 2014 20:03:53 GMT I am not sure what the 32 bit version of libgvfscommon0 is doing: it seems to be left over from Mageia 4. I will attach boot.txt and log.txt
Created attachment 11239 [details] log file
Created attachment 11240 [details] log file log.txt.gz
Created attachment 11241 [details] log file log.txt.gz Previous attachment was defined as text, but then I had to gzip it and forgot to change the type.
Attachment 11240 is obsolete: 0 => 1
(In reply to Martin Ward from comment #2) > > rpm -qa --last | grep x86_64 | wc -l > > returns: 3148 > > rpm -qa --last | grep i586 | wc -l > > returns: 284 > (In reply to Martin Ward from comment #5) > Created attachment 11241 [details] > log file log.txt.gz > I see hundreds of 32bit libraries on your 64bit system, some of which are devel packages and one normal 32bit package (remmina-plugins-vnc, for which a 64bit version exists). I've seen several users with upgrade problems because they had a mixed arch system. I'd advise to replace those packages with their 64bit versions. You'll know that "lib" in the package name is "lib64" for the 64bit version. Please report back whether that fixes the problem. It is probably best to disable the 32bit repositories.
I have uninstalled all the 32bit packages. This makes no difference to the boot. I also restored the nvidia driver and now get the same result as before: a black screen with no cursor.
(In reply to Martin Ward from comment #7) > I have uninstalled all the 32bit packages. > This makes no difference to the boot. > > I also restored the nvidia driver and now get the same result as before: > a black screen with no cursor. Thanks for having tried, Martin. Can you please, again (in case something changed), when you get the black screen, press "Ctrl Alt F2" simultaneously to get to a txt login prompt, then please become root and again run: journalctl -ab > boot.txt and attach it (you can obsolete the other files) CC'ing our kernel & drivers maintainers in case this issue is driver related and also our KDE team, in case this is an issue with SDDM
CC: (none) => kde, kernel
Attachment 11241 is obsolete: 0 => 1
Attachment 11239 is obsolete: 0 => 1
Created attachment 11249 [details] Output of journalctl -ab > boot.txt
(In reply to Martin Ward from comment #9) > Created attachment 11249 [details] > Output of journalctl -ab > boot.txt There are some lines that look weird to me when SDDM is started, but this looks worse: Aug 12 09:11:57 pingu6 kernel: Your BIOS is broken; DMAR reported at address 0! BIOS vendor: American Megatrends Inc.; Ver: F7 UA; Product Version: To be filled by O.E.M. Aug 12 09:11:57 pingu6 kernel: ------------[ cut here ]------------ Aug 12 09:11:57 pingu6 kernel: WARNING: CPU: 0 PID: 0 at drivers/iommu/dmar.c:845 warn_invalid_dmar.part.12+0x57/0x60 Aug 12 09:11:57 pingu6 kernel: Modules linked in: Aug 12 09:11:57 pingu6 kernel: CPU: 0 PID: 0 Comm: swapper Not tainted 4.14.131-desktop-1.mga6 #1 Aug 12 09:11:57 pingu6 kernel: Hardware name: Gigabyte Technology Co., Ltd. Z87X-UD4H/Z87X-UD4H-CF, BIOS F7 UA 08/12/2013 Aug 12 09:11:57 pingu6 kernel: task: ffffffff950134c0 task.stack: ffffffff95000000 Aug 12 09:11:57 pingu6 kernel: RIP: 0010:warn_invalid_dmar.part.12+0x57/0x60 Aug 12 09:11:57 pingu6 kernel: RSP: 0000:ffffffff95003e08 EFLAGS: 00010086 Aug 12 09:11:57 pingu6 kernel: RAX: 000000000000008b RBX: 0000000000000000 RCX: ffffffff9504f5a8 Aug 12 09:11:57 pingu6 kernel: RDX: 0000000000000000 RSI: 0000000000000082 RDI: 0000000000000002 Aug 12 09:11:57 pingu6 kernel: RBP: ffffffff94e585d2 R08: 0000000000000001 R09: 00000000000000a1 Aug 12 09:11:57 pingu6 kernel: R10: ffffffff95003d80 R11: 00000000000000a1 R12: ffffffff9569201c Aug 12 09:11:57 pingu6 kernel: R13: ffffffff9569205c R14: 0000000000000000 R15: 00000000ca44e018 Aug 12 09:11:57 pingu6 kernel: FS: 0000000000000000(0000) GS:ffff8b2b6ec00000(0000) knlGS:0000000000000000 Aug 12 09:11:57 pingu6 kernel: CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 Aug 12 09:11:57 pingu6 kernel: CR2: ffff8b289d693000 CR3: 000000055d00a001 CR4: 00000000000606b0 Aug 12 09:11:57 pingu6 kernel: Call Trace: Aug 12 09:11:57 pingu6 kernel: dmar_validate_one_drhd+0x99/0xd0 Aug 12 09:11:57 pingu6 kernel: dmar_walk_remapping_entries+0x99/0x1c0 Aug 12 09:11:57 pingu6 kernel: detect_intel_iommu+0x62/0xea Aug 12 09:11:57 pingu6 kernel: ? xen_swiotlb_init+0x4d0/0x4d0 Aug 12 09:11:57 pingu6 kernel: pci_iommu_alloc+0x51/0x6c Aug 12 09:11:57 pingu6 kernel: mem_init+0x16/0x98 Aug 12 09:11:57 pingu6 kernel: start_kernel+0x2a0/0x4df Aug 12 09:11:57 pingu6 kernel: secondary_startup_64+0xa5/0xb0 Aug 12 09:11:57 pingu6 kernel: Code: d9 0f 00 bf 01 00 00 00 49 89 c4 e8 14 d9 0f 00 4d 89 e9 4d 89 e0 48 89 c1 48 89 ea 48 89 de 48 c7 c7 d8 f1 ea 94 e8 14 e0 af ff <0f> 0b 5b 5d 41 5c 41 5d c3 e8 3b 04 23 00 85 d2 7e 61 53 b8 01 Aug 12 09:11:57 pingu6 kernel: ---[ end trace f2dc2a78a3696619 ]--- Assigning to the kernel maintainers for a correct interpretation of the above. (Note that there was also: Aug 12 09:12:08 pingu6 modprobe[2325]: modprobe: ERROR: missing parameters. See -h.)
Assignee: bugsquad => kernel
Hi, thanks for reporting this bug. We are sorry, but we no longer maintains this version of Mageia. Please upgrade to the latest version and reopen this bug against that version if this bug exists there. As a result we are setting this bug to CLOSED:OLD
Status: NEW => RESOLVEDResolution: (none) => OLDCC: (none) => ouaurelien