Bug 15109 - from lilo it can not boot, from grub it is booting and shuts down the machine
Summary: from lilo it can not boot, from grub it is booting and shuts down the machine
Status: RESOLVED OLD
Alias: None
Product: Mageia
Classification: Unclassified
Component: Release (media or process) (show other bugs)
Version: 5
Hardware: x86_64 Linux
Priority: Normal normal
Target Milestone: ---
Assignee: All Packagers
QA Contact:
URL:
Whiteboard: MGA5TOO
Keywords:
Depends on:
Blocks:
 
Reported: 2015-01-22 05:30 CET by Vuk Vujovic
Modified: 2018-06-26 11:41 CEST (History)
10 users (show)

See Also:
Source RPM: lilo
CVE:
Status comment:


Attachments

Description Vuk Vujovic 2015-01-22 05:30:24 CET
Description of problem:


Version-Release number of selected component (if applicable):

Kernel 3.18.3-server

How reproducible:

Installation of kernel and all related RPMS and lilo

Steps to Reproduce:
1. Install kernel and all related RPMS to it on machine
2. Let lilo configure itself for newest kernel
3. Reboot machine and choose that kernel
4. You will get (at least I got) error message: "Kernel and initrd memory conflict"



Reproducible: 

Steps to Reproduce:
claire robinson 2015-01-22 10:53:01 CET

CC: (none) => mageia, pterjan, tmb

Comment 1 claire robinson 2015-01-22 11:26:19 CET
Is it the same as bug 9848 Vuk?
Comment 2 claire robinson 2015-01-22 11:28:04 CET
hmm perhaps not
Comment 3 Vuk Vujovic 2015-01-23 23:31:30 CET
(In reply to claire robinson from comment #2)
> hmm perhaps not

I have this reports. Installed all files from repo as update to Cauldron and received 3.18.3-server-2 kernel. Kernel installed in Lilo and still getting the same error on boot (Kernel and Initrd memory conflict). However 3.18.3-desktop-2 kernel booted ok (not including shutdown of machine during boot which is I think my hardware failure).

So 3.18.3-server-2 can not boot because initrd is somewhere wrong and 3.18.3-desktop-2 boots without problem and I'm now using it. 

I noticed that machine is slightly faster or faster when using desktop kernel instead of server.

Vuk Vujovic
Samuel Verschelde 2015-05-31 23:33:46 CEST

Source RPM: kernel-server-3.18.3 => kernel-server-3.18.3 lilo
Whiteboard: (none) => MGA5TOO

Comment 4 alan headrick 2015-12-05 14:57:48 CET
I have been trying to upgrade/update from mga4 to 5 the last week and get the same error listed above. Also using lilo. It seems mga5 is installed, but I have to manually select one of the older mga4 kernels to boot. It was suggested to me to post this here. Hopefully this is OK.
Thanks duder

CC: (none) => 1alan.headrick, sysadmin-bugs
Component: RPM Packages => Release (media or process)
Version: Cauldron => 5
Source RPM: kernel-server-3.18.3 lilo => desktop-4.1.13-2.mga5

Comment 5 Luke Jones 2015-12-05 19:57:17 CET
Has anyone tried with grub2 installed? Would like to rule out the boot loader.

Of note about lilo, it's last code commit was well over 2 years ago and many distributions are now dropping support for it. Also, grub (not grub2) is no longer in active development due to grub2 being the replacement.
I may file a report about these.

CC: (none) => luke.nukem.jones

Comment 6 Florian Hubold 2016-01-30 08:55:06 CET
Also reported via https://forums.mageia.org/en/viewtopic.php?f=7&t=10498 although I don't understand the actual issue, kernel and initrd images look just fine.

CC: (none) => doktor5000

Comment 7 Thomas Backlund 2016-01-30 13:18:00 CET
Its not the kernel or initrd as such that is the problem, it's when lilo tries to map kernel and initrd to memory it gets in trouble...
probably because kernel size has grown...

I havent looked at lilo lately to see if it can be fixed for mga5... for mga6 we should simply drop lilo...

as for grub legacy I guess it will live for a long time as it is a small nice loader with simple clean configs, and does not have any of the useless bloat that grub2 comes whit and its uggly configs.

The only thing we wont patch grub legacy for is efi support, so as times goes by...

Then again we could look at other clean efi loaders, such as gummiboot as default...
Comment 8 Florian Hubold 2016-01-30 15:02:05 CET
(In reply to Thomas Backlund from comment #7)
> Its not the kernel or initrd as such that is the problem, it's when lilo
> tries to map kernel and initrd to memory it gets in trouble...
> probably because kernel size has grown...

Yeah also read similar things after posting here. From what I found, in theory there should be no size limit to either kernel or initrd that lilo handles.
But this only seems to happen on upgrades, found the same issue "Kernel and initrd memory conflict" reported for lilo for a MDK 9.0 > 9.2 upgrade.


This seems to be the problem code: https://github.com/a2o/lilo/blob/master/src/second.S#L1862

I've also asked the guy in the forum to try to switch to grub/grub2 just to see if that fixes this issue.
Comment 9 Samuel Verschelde 2016-11-01 12:41:57 CET
Assigning to all packagers collectively since lilo has no registered maintainer for now. But also CCing kernel maintainers.

CC: (none) => kernel
Assignee: bugsquad => pkg-bugs
Source RPM: desktop-4.1.13-2.mga5 => lilo

Comment 10 Thierry Vignaud 2016-11-09 17:58:11 CET
I've seen such similar behaviour in the past (it was reboot not poweroff), around a year ago but I don't remember what was the exact issue.

CC: (none) => thierry.vignaud

Comment 11 Marja Van Waes 2018-06-26 11:41:00 CEST
Hi Vuk,

Thank you for having taken the needed time to report this issue!

We regret if this issue didn't get fixed in Mageia 5.

Mageia 5 has officially reached its End of Life on December 31st, 2017 https://blog.mageia.org/en/2017/11/07/mageia-5-eol-postponed/
It only continued to get important security updates since then, but non-security bugs have no chance of still getting fixed.

I haven't seen anyone report this issue in Mageia 6 or later (and even this bug report for Mageia 5 had no action in a very long time.

Closing as OLD.

Please reopen this report and change its "Version:" at the top left to "6", if the same bug still exists in Mageia 6.

Cheers,
Marja

Resolution: (none) => OLD
Status: NEW => RESOLVED
CC: (none) => marja11


Note You need to log in before you can comment on or make changes to this bug.