Cinnamon segfault at 28 ip 00007fb707b32ae2 sp 00007ffd444e3780 error 6 in libmuffin.so.0.0.0[7fb707afe000+74000]
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
muffin (Ubuntu) |
Fix Released
|
Medium
|
Joshua Peisach | ||
Focal |
Fix Released
|
Undecided
|
Joshua Peisach |
Bug Description
This is an upstream issue-read the following bug reports.
https:/
https:/
An SRU patch is now available.
This is a REGRESSION of a commit in Cinnamon 4.0.0 (https:/
Debian Bug Patch: https:/
SRU TEMPLATE:
[Impact]
* In fallback mode Cinnamon/Muffin segfaults and makes the desktop unusable.
* This would occur by having two monitors-a laptop and an external with the laptop off and the external on-supposedly with NVIDIA as a driver.
[Test Case]
* Upstream commits example to replicate bug:
"Example to reproduce: Laptop monitor + external monitor. Arrange laptop
monitor to the right of the external one. Using nvidia-450 (confirmed
for me, at least), turn laptop monitor off. Ignoring horrible lag from
driver, turn the laptop monitor back on (arranging it still to the right
of the external monitor)."
[Regression Potential]
* Regressions manifesting is rather barely likely. Latest versions of Muffin patches this and so is upstream unstable have this patched. Regressions shouldn't be a threat unless:
1) Dependency problem
2) New versions of C causing incompatability problems or versions of Gobject-
* In the event of a regression, which is unlikely as of Aug 13 (UTC), as Muffin's update would've been released a while ago (24 days), 2 days ago the version was bumped. A new upstream version would likely patch this problem aswell.
* After 24 days it seems there is nothing to worry about regarding regressions but of course, always something to worry about.
* It's been about enough time, nearly a month to spot any regressions from this fix.
* This is actually a SOLVE of a regression
* This regression fix goes back to Cinnamon 4.0.0-or when Eoan Ermine was released by getting refresh rates form xrandr
* In upstream Mint this was an SRU/release as seen here: https:/
[Other Info]
* I couldn't pbuilder-dist for some reason. Missing dependencies or stuff didn't exist-in sbuild it did though and my debuild -S did.
* I will work on patching this in Groovy through debian unstable and will attach the bug report for groovy as part of this situation/
Video: https:/
[Test Case #1]
Dell Inspiron 15-3567, Intel Core i5 7th Gen, for external monitor/TV used Samsung 1920x TV.
1) Start up laptop and login with Cinnamon
2) Plug in TV (external monitor), HDMI cable
3) Open display cinnamon settings module
4) Arrange the laptop monitor ON THE RIGHT OF THE EXTERNAL MONITOR
5) Set the TV as the primary display.
6) Close the laptop lid. This will turn off the laptop monitor.
7) Open some windows, play around with the TV monitor for a bit.
8) Reopen the laptop monitor.
9) Graphics will lag and cause black borders and screen tearing. In this case it was on the laptop monitor but some other times when I tried it happened on the TV.
[Test Case #1 Notes]
This is WITHOUT THE PATCH.
I noticed that you dont even have to turn off and re-turn on the laptop monitor for the screen tearing/black borders to occur.
Other times the laptop monitor would just turn off and you wouldn't be able to turn it back on unless you forced the Display settings module to make the monitor active.
Apport or Ubuntu-bug didn't popup a window showing of a crash.
Black borders at 2:15 in video.
[TEST CASE #2]
10) Software and Updates -> Development -> Enable -proposed repos
11) Refresh cache
12) Open terminal
13) sudo apt update, sudo apt upgrade
14) Reboot
15) Login and try above bug replication again
16) No errors/bugs.
VERIFICATION DONE
Patch does work-without the patch with the dual-monitor you would get instant black bordering. This patch removes it.
I emailed Michael Webster from upstream about the patch. This was his response:
"
- Immediately below the code insertion is a check that info != NULL, so the code prior to that commit allowed that info could be NULL (this code was existing from our original commit when we forked mutter ~9 years ago).
- The crash was specifically due to attempting to dereference a NULL info.
"
He hasn't tested previous versions but I have confirmed 4.8, the only one Ubuntu is maintaining for now and the regression was introduced in 4.0 (Eoan).
ProblemType: Bug
DistroRelease: Ubuntu 20.04
Package: muffin 4.4.3-1
ProcVersionSign
Uname: Linux 5.4.0-40-generic x86_64
ApportVersion: 2.20.11-0ubuntu27.3
Architecture: amd64
CasperMD5CheckR
CurrentDesktop: ubuntu:GNOME
Date: Sun Jul 19 21:49:09 2020
InstallationDate: Installed on 2020-04-23 (87 days ago)
InstallationMedia: Ubuntu 20.04 LTS "Focal Fossa" - Release amd64 (20200423)
ProcEnviron:
TERM=xterm-
PATH=(custom, no user)
XDG_RUNTIME_
LANG=en_US.UTF-8
SHELL=/bin/bash
SourcePackage: muffin
UpgradeStatus: No upgrade log present (probably fresh install)
Changed in muffin (Ubuntu): | |
importance: | Undecided → Medium |
tags: | added: sru |
Changed in muffin (Ubuntu Focal): | |
assignee: | nobody → Joshua Peisach (itzswirlz) |
description: | updated |
description: | updated |
The attachment "muffin_ 4.4.3-1ubuntu0. 1.debdiff" seems to be a debdiff. The ubuntu-sponsors team has been subscribed to the bug report so that they can review and hopefully sponsor the debdiff. If the attachment isn't a patch, please remove the "patch" flag from the attachment, remove the "patch" tag, and if you are member of the ~ubuntu-sponsors, unsubscribe the team.
[This is an automated message performed by a Launchpad user owned by ~brian-murray, for any issue please contact him.]