Description of problem: During install near the end the boot loader installation has failed and the error was displayed in a window but then I was not given a second chance to select a different boot loader after that. Instead, the installer just continued and said "Congratulations" and wanted to reboot into the non working boot loader it just has failed to set up. Version-Release number of selected component (if applicable): 4.1 boot-nonfree.iso x86_64 but probably not specific to this iso How reproducible: I was installing in a virtual machine with virtio harddisk and selected lilo text mode as the boot loader (why it failed is probably a separate bug that I don't report now, feel free to add a bug for that if needed). Reproducible: Steps to Reproduce:
well, we needs the logs in order to understand what happened. Can you retrive /root/drakx/ddebug.log? See libguestfs-tools
Keywords: (none) => NEEDINFOCC: (none) => thierry.vignaud
This bug is not about why the installation of the boot loader has failed so it is not important what happened. What's important is that if this step fails the installer should not continue as if it was successful but go back to the summary screen so I can select another bootloader and retry with that. It should only continue and reboot if the bootloader has successfully installed otherwise the user is left with a non working system. (I was able to fix the installation from the rescue system but less expirenced users might give up if this happens or have to reinstall to try a different bootloader.) Anyway, if you want to also fix the particular case with lilo and virtio here are the details: * fs::get::device2part: unknown device <</dev/vda>> * running: keytab-lilo.pl us with root /mnt * program not found: keytab-lilo.pl * writing lilo config to /mnt/etc/lilo.conf * Installing boot loader on /dev/vda... * running: echo | lilo with root /mnt * error: : Fatal: Linux experimental device 0xfc00 needs to be defined. Check 'man lilo.conf' under 'disk=' and 'max-partitions=' matchbox-wm: X error warning (0xa13ced): 140 (opcode: 138) * step "summary" took: 0:03:08 * step `summary' finished I had to manually add this to lilo.conf and reinstall lilo to fix the installation: disk=/dev/vda bios=0x80 max-partitions=7 But again, this is really a separate bug and fixing it so lilo works would not fix the general problem in case the bootloader installation fails for some other reason.
I still need the full logs in order to understand what happened and thus what is the bug. Please attach it (do not paste it)
Created attachment 5222 [details] ddebug.log Attached ddebug.log
Needinfo no longer applies.
Keywords: NEEDINFO => (none)Assignee: bugsquad => thierry.vignaud
Setting version to cauldron since we can't fix Mageia 4 installers.
Version: 4 => CauldronSource RPM: Installer in 4.1 boot-nonfree.iso => (none)Whiteboard: (none) => MGA4TOO
Target Milestone: --- => Mageia 6Whiteboard: MGA4TOO => (none)
Lilo has issues with /dev/vda. We cannot fix a released installer and we will only support grub2 in mga6. The thing to do in such a case is to rerun the installer & choose a different bootloader when using mga4
Status: NEW => RESOLVEDResolution: (none) => WONTFIXSummary: Installer continues after failed boot loader installation => Installer continues after failed boot loader installation (LILO: "Fatal: Linux experimental device 0xfc00 needs to be defined")