Bug 19771 - 6_rc0: sddm segfaults in vendor VirtualBox after 4.8.7-desktop-1 kernel update
Summary: 6_rc0: sddm segfaults in vendor VirtualBox after 4.8.7-desktop-1 kernel update
Status: RESOLVED FIXED
Alias: None
Product: Mageia
Classification: Unclassified
Component: RPM Packages (show other bugs)
Version: Cauldron
Hardware: x86_64 Linux
Priority: Normal critical
Target Milestone: ---
Assignee: KDE maintainers
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks: 19782
  Show dependency treegraph
 
Reported: 2016-11-13 19:23 CET by Bit Twister
Modified: 2016-11-26 16:38 CET (History)
2 users (show)

See Also:
Source RPM: sddm-0.14.0-10.mga6.src.rpm
CVE:
Status comment:


Attachments
journalctl -b | grep sddm (10.48 KB, text/plain)
2016-11-13 19:24 CET, Bit Twister
Details

Description Bit Twister 2016-11-13 19:23:01 CET
Description of problem:  6_rc0:

 sddm segfaults in vendor VirtualBox after 4.8.7-desktop-1 kernel update


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


How reproducible: always


Steps to Reproduce:
1. urpmi --download-all  --downloader wget --auto --auto-update


Does not mater if I add nomodeset or iomem=relaxed or vboxvideo.modeset=0
sddm still segfaulted.
Comment 1 Bit Twister 2016-11-13 19:24:38 CET
Created attachment 8657 [details]
journalctl -b | grep sddm
Comment 2 Marja Van Waes 2016-11-13 21:03:44 CET
(In reply to Bit Twister from comment #0)
> Description of problem:  6_rc0:
> 
>  sddm segfaults in vendor VirtualBox after 4.8.7-desktop-1 kernel update
> 
>

What else was then updated? Kernel updates never come alone ;-)

(run e.g. "rpm -qa --last | head -n50 > last50.txt" or check the journalctl logs from around the kernel update)

Does the problem get fixed if you run updates again? 

If not, do you happen to have an unhappy mirror?
http://mirrors.mageia.org/status

CC: (none) => kernel, marja11
Assignee: bugsquad => kde

Comment 3 Bit Twister 2016-11-13 21:20:45 CET
(In reply to Marja van Waes from comment #2)

> What else was then updated? Kernel updates never come alone ;-)

chromium-browser-54.0.2840.100-1.mga6.x86_64.rpm
chromium-browser-stable-54.0.2840.100-1.mga6.x86_64.rpm
cpupower-4.8.7-1.mga6.x86_64.rpm
drakx-kbd-mouse-x11-1.18-1.mga6.x86_64.rpm
drakx-kbd-mouse-x11-text-1.18-1.mga6.x86_64.rpm
hexchat-2.12.3-1.mga6.x86_64.rpm
kernel-desktop-4.8.7-1.mga6-1-1.mga6.x86_64.rpm
kernel-desktop-devel-4.8.7-1.mga6-1-1.mga6.x86_64.rpm
kernel-desktop-devel-latest-4.8.7-1.mga6.x86_64.rpm
kernel-desktop-latest-4.8.7-1.mga6.x86_64.rpm
kernel-doc-4.8.7-1.mga6.noarch.rpm
kernel-source-4.8.7-1.mga6-1-1.mga6.noarch.rpm
kernel-source-latest-4.8.7-1.mga6.noarch.rpm
lib64git2_24-0.24.1-2.mga6.x86_64.rpm
lib64rpm7-4.13.0-6.mga6.x86_64.rpm
lib64rpmbuild7-4.13.0-6.mga6.x86_64.rpm
libdrakx-kbd-mouse-x11-1.18-1.mga6.x86_64.rpm
python2-rpm-4.13.0-6.mga6.x86_64.rpm
rpm-4.13.0-6.mga6.x86_64.rpm
rpm-build-4.13.0-6.mga6.x86_64.rpm
x11-driver-video-intel-2.99.917-30.mga6.x86_64.rpm


> Does the problem get fixed if you run updates again? 

Nope.


> If not, do you happen to have an unhappy mirror?

No way, Methodology is pull_updates on node wb.
ssh tb, pull_updates
pull_updates rsyncs rpms from wb.
On TestBed (tb) install_updates
reboot.

Ok, launch Virtualbox on wb and then vb.
pull_updates, rsyncs wb rpms to vb.
install_updates and I reboot and it fails.

On wb, I run install_updates, reboot and all works.
Same set of rpms used where applicable.
Bit Twister 2016-11-15 10:44:02 CET

Severity: normal => critical

Comment 4 Bit Twister 2016-11-21 00:23:58 CET
$ uname -r
4.8.9-desktop-1.mga6

Workaround: add nomodeset to kernel line and remove any vga=xxx argument.
Bit Twister 2016-11-25 04:55:19 CET

Blocks: (none) => 19782

Comment 5 Bit Twister 2016-11-25 21:45:54 CET
removed nomodeset and added vga=791 back to kernel boot line.
sddm no longer segfaulta and I do have a working display in runlevel 5.
Comment 6 Bit Twister 2016-11-25 21:46:26 CET
Sorry forgot to mark this resolved.

Status: NEW => RESOLVED
Resolution: (none) => FIXED

Comment 7 Samuel Verschelde 2016-11-26 11:58:26 CET
Even if there's a workaround, it's still segfaulting in the other case so the bug is not fixed.

Status: RESOLVED => REOPENED
Resolution: FIXED => (none)

Comment 8 Bit Twister 2016-11-26 16:07:09 CET
(In reply to Samuel Verschelde from comment #7)
> Even if there's a workaround, it's still segfaulting in the other case so
> the bug is not fixed.

You need to provide the "other case" so we know what you are talking about.

All suggested kernel arguments to date did not work until I removed vga=xxx.

I provide vga=xxx on all my installs host and guest alike. With the latest sddm update I no longer added any of the suggested arguments and vga=792 went back to working in my vb guests.

Current grub2 kernel arguments are built from
   $ grep -i cmd /etc/default/grub
   GRUB_CMDLINE_LINUX_DEFAULT=" noiswmd"
   GRUB_CMDLINE_LINUX="ipv6.disable=1 audit=0 vga=792"


GRUB_CMDLINE_LINUX= line is generated by my install script.
Comment 9 Samuel Verschelde 2016-11-26 16:38:17 CET
Maybe I misread your comments. I though you found a workaround but that the initial segfault was still there. It's apparently not true, so closing again.

Status: REOPENED => RESOLVED
Resolution: (none) => FIXED


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