[nvidia] Corrupted wallpaper after resuming from suspend or hibernation

Bug #1809407 reported by Kevin McMurtrie
512
This bug affects 95 people
Affects Status Importance Assigned to Milestone
GNOME Shell
Fix Released
Unknown
mutter (Ubuntu)
Fix Released
High
Daniel van Vugt
Disco
Won't Fix
Undecided
Daniel van Vugt
Eoan
Fix Released
High
Daniel van Vugt

Bug Description

[Impact]

The Nvidia (Xorg) driver corrupts the desktop wallpaper (by design (1)) upon resuming from suspend. This creates a poor user experience and gives the impression of severe memory corruption.

(1) https://www.khronos.org/registry/OpenGL/extensions/NV/NV_robustness_video_memory_purge.txt

[Test Case]

0. Set up an Nvidia-only system using the proprietary Nvidia Xorg driver.
1. Suspend and resume the machine.
Expect: The wallpaper still looks like it did before suspending.

[Regression Potential]

Low. The exact same patch is in Gnome 3.34 (Ubuntu 19.10) already and has been used for a couple of months without issue.

[Original Report]

https://trello.com/c/i5hENxSO

Every other wake from sleep presents a corrupted screen with no unlock app. There's a dock and system menu on a black background with pixel garbage. Some personal information in the dock's large window thumbnails. Bitmap garbage flashes on the screen when interacting with anything. Only workaround is doing another sleep-wake cycle.

ProblemType: Bug
DistroRelease: Ubuntu 18.10
Package: xorg 1:7.7+19ubuntu8
ProcVersionSignature: Ubuntu 4.18.0-12.13-generic 4.18.17
Uname: Linux 4.18.0-12-generic x86_64
NonfreeKernelModules: nvidia_modeset nvidia
.proc.driver.nvidia.gpus.0000.4b.00.0: Error: [Errno 21] Is a directory: '/proc/driver/nvidia/gpus/0000:4b:00.0'
.proc.driver.nvidia.registry: Binary: ""
.proc.driver.nvidia.version:
 NVRM version: NVIDIA UNIX x86_64 Kernel Module 390.87 Tue Aug 21 12:33:05 PDT 2018
 GCC version: gcc version 8.2.0 (Ubuntu 8.2.0-7ubuntu1)
ApportVersion: 2.20.10-0ubuntu13.1
Architecture: amd64
BootLog: Error: [Errno 13] Permission denied: '/var/log/boot.log'
CompizPlugins: No value set for `/apps/compiz-1/general/screen0/options/active_plugins'
CompositorRunning: None
CurrentDesktop: ubuntu:GNOME
Date: Thu Dec 20 22:19:29 2018
DistUpgraded: 2018-12-01 23:07:50,385 ERROR got error from PostInstallScript ./xorg_fix_proprietary.py (g-exec-error-quark: Failed to execute child process “./xorg_fix_proprietary.py” (No such file or directory) (8))
DistroCodename: cosmic
DistroVariant: ubuntu
DkmsStatus:
 nvidia, 390.87, 4.18.0-11-generic, x86_64: installed
 nvidia, 390.87, 4.18.0-12-generic, x86_64: installed
 virtualbox, 5.2.18, 4.18.0-11-generic, x86_64: installed
 virtualbox, 5.2.18, 4.18.0-12-generic, x86_64: installed
ExtraDebuggingInterest: Yes, if not too technical
GraphicsCard:
 NVIDIA Corporation GP106 [GeForce GTX 1060 3GB] [10de:1c02] (rev a1) (prog-if 00 [VGA controller])
   Subsystem: eVga.com. Corp. GP106 [GeForce GTX 1060 3GB] [3842:6162]
InstallationDate: Installed on 2018-10-28 (54 days ago)
InstallationMedia: Ubuntu 18.04.1 LTS "Bionic Beaver" - Release amd64 (20180725)
MachineType: Gigabyte Technology Co., Ltd. Default string
ProcEnviron:
 TERM=xterm-256color
 PATH=(custom, no user)
 XDG_RUNTIME_DIR=<set>
 LANG=en_US.UTF-8
 SHELL=/bin/bash
ProcKernelCmdLine: BOOT_IMAGE=/boot/vmlinuz-4.18.0-12-generic root=UUID=a8c839e7-c579-4699-999a-748f804ff9e0 ro quiet splash vt.handoff=1
SourcePackage: xorg
Symptom: display
UpgradeStatus: Upgraded to cosmic on 2018-12-02 (18 days ago)
dmi.bios.date: 06/08/2017
dmi.bios.vendor: American Megatrends Inc.
dmi.bios.version: F24
dmi.board.asset.tag: Default string
dmi.board.name: X99P-SLI-CF
dmi.board.vendor: Gigabyte Technology Co., Ltd.
dmi.board.version: x.x
dmi.chassis.asset.tag: Default string
dmi.chassis.type: 3
dmi.chassis.vendor: Default string
dmi.chassis.version: Default string
dmi.modalias: dmi:bvnAmericanMegatrendsInc.:bvrF24:bd06/08/2017:svnGigabyteTechnologyCo.,Ltd.:pnDefaultstring:pvrDefaultstring:rvnGigabyteTechnologyCo.,Ltd.:rnX99P-SLI-CF:rvrx.x:cvnDefaultstring:ct3:cvrDefaultstring:
dmi.product.family: Default string
dmi.product.name: Default string
dmi.product.sku: Default string
dmi.product.version: Default string
dmi.sys.vendor: Gigabyte Technology Co., Ltd.
version.compiz: compiz N/A
version.libdrm2: libdrm2 2.4.95-1
version.libgl1-mesa-dri: libgl1-mesa-dri 18.2.2-0ubuntu1
version.libgl1-mesa-glx: libgl1-mesa-glx 18.2.2-0ubuntu1
version.nvidia-graphics-drivers: nvidia-graphics-drivers-* N/A
version.xserver-xorg-core: xserver-xorg-core 2:1.20.1-3ubuntu2.1
version.xserver-xorg-input-evdev: xserver-xorg-input-evdev N/A
version.xserver-xorg-video-ati: xserver-xorg-video-ati 1:18.1.0-1
version.xserver-xorg-video-intel: xserver-xorg-video-intel 2:2.99.917+git20171229-1ubuntu1
version.xserver-xorg-video-nouveau: xserver-xorg-video-nouveau 1:1.0.15-3

Related branches

Revision history for this message
Kevin McMurtrie (kevinmcmurtrie) wrote :
Revision history for this message
Daniel van Vugt (vanvugt) wrote :

This sounds like a combination of bug 1769383 and something else. Can you please provide a photo of the issue so we can figure out what the second part is?

affects: xorg (Ubuntu) → gnome-shell (Ubuntu)
Changed in gnome-shell (Ubuntu):
status: New → Incomplete
tags: added: nvidia
Changed in gdm3 (Ubuntu):
status: New → Incomplete
summary: - No unlock screen
+ [nvidia] Lock screen is a black background with pixel garbage. Bitmap
+ garbage flashes on the screen when interacting with anything.
summary: - [nvidia] Lock screen is a black background with pixel garbage. Bitmap
- garbage flashes on the screen when interacting with anything.
+ [nvidia] After resume the unlock screen is a black background with pixel
+ garbage. Bitmap garbage flashes on the screen when interacting with
+ anything.
tags: added: resume suspend-resume
Revision history for this message
Kevin McMurtrie (kevinmcmurtrie) wrote : Re: [nvidia] After resume the unlock screen is a black background with pixel garbage. Bitmap garbage flashes on the screen when interacting with anything.

I did some random trial and error debugging.

There's always a login screen on console 1 and there's always a desktop GUI on console 2. It's sometimes waking from sleep into display 2 while still locked. There's a dock, a system menu, and random bitmap garbage. Hitting CTRL-ALT-F1 goes to the login screen. After logging in, it jumps to a normally working display 2.

Not sure if it's related: CTRL-ALT-fkey is sometimes generating app keyboard events rather than switching consoles.

Installing 18.04 completely corrupted my system and I had wipe the disk. This isn't an old crusty installation.

Revision history for this message
Launchpad Janitor (janitor) wrote :

[Expired for gdm3 (Ubuntu) because there has been no activity for 60 days.]

Changed in gdm3 (Ubuntu):
status: Incomplete → Expired
Revision history for this message
Launchpad Janitor (janitor) wrote :

[Expired for gnome-shell (Ubuntu) because there has been no activity for 60 days.]

Changed in gnome-shell (Ubuntu):
status: Incomplete → Expired
Changed in gdm3 (Ubuntu):
status: Expired → New
Changed in gnome-shell (Ubuntu):
status: Expired → New
Revision history for this message
Launchpad Janitor (janitor) wrote :

Status changed to 'Confirmed' because the bug affects multiple users.

Changed in gdm3 (Ubuntu):
status: New → Confirmed
Changed in gnome-shell (Ubuntu):
status: New → Confirmed
Changed in nvidia-graphics-drivers-390 (Ubuntu):
status: New → Confirmed
Changed in nvidia-graphics-drivers-418 (Ubuntu):
status: New → Confirmed
no longer affects: gdm3 (Ubuntu)
Changed in mutter (Ubuntu):
status: New → Confirmed
tags: added: disco
Revision history for this message
Luca Osvaldo Mastromatteo (lukycrociato) wrote :

Hi Daniel, I've seen you marked my bug as this duplicate, but for me this doesn't happen on Ubuntu 18.10, only after the 19.04 upgrade

Revision history for this message
Daniel van Vugt (vanvugt) wrote :

That's OK. This bug is tagged 'disco' meaning it also occurs in 19.04.

Revision history for this message
Luca Osvaldo Mastromatteo (lukycrociato) wrote :

Okay, it's fine anyway I tested 18.10 again as well with another distribution with roughly the same packages and I was not able to reproduce the issue

I'll try Debian testing

Revision history for this message
Luca Osvaldo Mastromatteo (lukycrociato) wrote :

Hi, in debian testing this does not happen, neither in debian unstable or other distros with the same gnome-shell package version

I confirm it's a regression from 18.10 to 19.04

In 18.10 it works fine

summary: - [nvidia] After resume the unlock screen is a black background with pixel
- garbage. Bitmap garbage flashes on the screen when interacting with
- anything.
+ [nvidia] Corrupted wallpaper after resuming from sleep
summary: - [nvidia] Corrupted wallpaper after resuming from sleep
+ [nvidia] Corrupted wallpaper after resuming from suspend
tags: added: xenial
Revision history for this message
Daniel van Vugt (vanvugt) wrote : Re: [nvidia] Corrupted wallpaper after resuming from suspend

Hmm, actually, if the problem did get worse or more common in 19.04 then we should use bug 1822179 instead.

Certainly after July [1], Ubuntu 18.10 won't be relevant any more.

[1] https://wiki.ubuntu.com/Releases

Revision history for this message
Ryan Ward (ryandward) wrote :

Just chiming in from Arch Linux, we got it too.

Revision history for this message
Jeremy Soller (jackpot51) wrote :

There is an upstream bug related to GNOME 3.32 here: https://gitlab.gnome.org/GNOME/gnome-shell/issues/1084

Revision history for this message
Jeremy Soller (jackpot51) wrote :

I was able to work around this with the attached patch. Apply by adding the file to debian/patches and its name to the end of debian/patches/series

Changed in gnome-shell:
status: Unknown → New
Revision history for this message
Ubuntu Foundations Team Bug Bot (crichton) wrote :

The attachment "nvidia-background-workaround.patch" seems to be a patch. If it isn't, please remove the "patch" flag from the attachment, remove the "patch" tag, and if you are a member of the ~ubuntu-reviewers, unsubscribe the team.

[This is an automated message performed by a Launchpad user owned by ~brian-murray, for any issues please contact him.]

tags: added: patch
Revision history for this message
Ghanshyam (gparvind) wrote :

I got this bug too after upgrading to 19.04 release with nvidia driver 418.
I see the corruption in lock screen and desktop background.

Additionally, if VS Code is open, terminal window in the vs code becomes blank. This used to happen in 18.10 as well. Not sure if it's related but just as additional info if it helps.

no longer affects: nvidia-graphics-drivers-390 (Ubuntu)
no longer affects: nvidia-graphics-drivers-418 (Ubuntu)
Changed in gnome-shell (Ubuntu):
importance: Undecided → High
Changed in mutter (Ubuntu):
importance: Undecided → High
Changed in gnome-shell (Ubuntu):
status: Confirmed → Triaged
Changed in mutter (Ubuntu):
status: Confirmed → Triaged
Revision history for this message
Daniel van Vugt (vanvugt) wrote :

Also in Trello: https://trello.com/c/i5hENxSO

Though all conversations really for now should occur upstream: https://gitlab.gnome.org/GNOME/gnome-shell/issues/1084

description: updated
tags: added: eoan
tags: added: rls-ee-incoming
Revision history for this message
Gwolf2u (gwolf2u-website) wrote :

What I've made a temp fix is this
place file in /lib/systemd/system-sleep
credits to: Paul Knopf (https://gitlab.gnome.org/GNOME/gnome-shell/issues/1084)

file content:
#!/bin/bash
#place file in /lib/systemd/system-sleep

PROGNAME=$(basename "$0")
state=$1
action=$2

function log {
    logger -i -t "$PROGNAME" "$*"
}

log "Running $action $state"

if [[ $state == post ]]; then
    log "WAKE UP"
    #gsettings set org.gnome.desktop.background picture-uri "file:///usr/share/backgrounds/Xplo_by_Hugo_Cliff.png"
    dbus-send --type=method_call --dest=org.gnome.Shell /org/gnome/Shell org.gnome.Shell.Eval "string:global.reexec_self()"
fi

Revision history for this message
Stuardo -StR- Rodríguez (stuardo) wrote :

I added @gwolf2u-website patch as /lib/systemd/system-sleep/nvidia.sh and changed permissions to a+w. The problem persists

If I run `dbus-send --type=method_call --dest=org.gnome.Shell /org/gnome/Shell org.gnome.Shell.Eval "string:global.reexec_self()"` manually, gnome restarts (faster) but all my windows are shown in the same virtual-desktop. If I switch screens (ctr+alt+up or down), windows return to their original desktop

Revision history for this message
Stuardo -StR- Rodríguez (stuardo) wrote :

Sorry... chmod a+x (I haven't got my coffee yet)

Revision history for this message
Stuardo -StR- Rodríguez (stuardo) wrote :

I've just tested adding the

`dbus-send --type=method_call --dest=org.gnome.Shell /org/gnome/Shell org.gnome.Shell.Eval "string:global.reexec_self()"`

command to the /lib/systemd/system-sleep/nvidia.sh file, and it didn't work neither.

Revision history for this message
Sebastien Bacher (seb128) wrote :

Daniel, you seem to have a looked at that/pinged Jonas, I'm assigning to you

Changed in gnome-shell (Ubuntu):
assignee: nobody → Daniel van Vugt (vanvugt)
Changed in mutter (Ubuntu):
assignee: nobody → Daniel van Vugt (vanvugt)
Changed in gnome-shell (Ubuntu Disco):
assignee: nobody → Daniel van Vugt (vanvugt)
Changed in mutter (Ubuntu Disco):
assignee: nobody → Daniel van Vugt (vanvugt)
tags: removed: rls-ee-incoming
Revision history for this message
Launchpad Janitor (janitor) wrote :

Status changed to 'Confirmed' because the bug affects multiple users.

Changed in gnome-shell (Ubuntu Disco):
status: New → Confirmed
Changed in mutter (Ubuntu Disco):
status: New → Confirmed
Revision history for this message
Petter Sundlöf (petter-sundlof) wrote :

Tried to resume, and was greeted with this (attached photo).

It then threw me back to the display manager (lightdm?) where I had to enter password, then the mutter/window manager's unlock screen opened and I had to enter password again. No crash this time but very odd behaviour

Revision history for this message
Launchpad Janitor (janitor) wrote :

Status changed to 'Confirmed' because the bug affects multiple users.

Changed in gnome-shell (Ubuntu Bionic):
status: New → Confirmed
Changed in mutter (Ubuntu Bionic):
status: New → Confirmed
Revision history for this message
Petter Sundlöf (petter-sundlof) wrote :

Here's a video of a typical loop. It's still sometimes possible to type my password and press Enter to get out of this.

This is after a wake from sleep.

Revision history for this message
Daniel van Vugt (vanvugt) wrote :

Petter,

Sorry, I may have prematurely marked bug 1828358 as a duplicate of this one. It's not any more. Please continue using bug 1828358 instead.

Changed in gnome-shell (Ubuntu Bionic):
assignee: nobody → Daniel van Vugt (vanvugt)
Changed in mutter (Ubuntu Bionic):
assignee: nobody → Daniel van Vugt (vanvugt)
Changed in mutter (Ubuntu Eoan):
status: Triaged → In Progress
Revision history for this message
kanna kamui (kannalo) wrote :

I dont think this issue is exclusive to ubuntu I get the same thing on arch using gnome https://imgur.com/a/ILOLikj

Revision history for this message
Daniel van Vugt (vanvugt) wrote :

It's a general problem with the Nvidia driver. A fix is on the way here:
https://gitlab.gnome.org/GNOME/mutter/merge_requests/600

no longer affects: gnome-shell (Ubuntu)
no longer affects: gnome-shell (Ubuntu Bionic)
no longer affects: gnome-shell (Ubuntu Disco)
no longer affects: gnome-shell (Ubuntu Eoan)
Revision history for this message
Daniel van Vugt (vanvugt) wrote :

Since this is a potentially large set of fixes required (in the long term), I am explicitly only aiming to fix the background wallpaper corruption in this bug. If anyone finds other things corrupt upon resuming from suspend then please open additional bugs for those.

Revision history for this message
Daniel van Vugt (vanvugt) wrote :
Revision history for this message
Willian Tenfen Wazilewski (tw-willian) wrote :

Will this be fixed for disco as well?

Revision history for this message
Removed by request (removed7361641) wrote :

I'm also interested in this, will it be fixed for 19.04 too?

Revision history for this message
Daniel van Vugt (vanvugt) wrote :

The Ubuntu release process requires that we always get the fix into the _next_ release first. That means 19.10. After that we are free to propose it for 19.04 and earlier. I have nominated 19.10, 19.04 and 18.04 at the top of this bug.

summary: - [nvidia] Corrupted wallpaper after resuming from suspend
+ [nvidia] Corrupted wallpaper after resuming from suspend or hibernation
Revision history for this message
satmandu (satadru-umich) wrote :

I fixed the workaround script mentioned above here:

https://gist.github.com/satmandu/2f2da70a7a6bbbacdba01ef8fb8c18f4

This "works" on my system.

(dbus-send doesn't get the right info to work unless you give it the info of the logged in user, which "machinectl shell" can do.)

Hopefully the fix gets merged soon and I can stop using this ugly hack. :)

#!/bin/bash
# Place file in /lib/systemd/system-sleep/
# Run chmod +x on the file.
# This works around the wallpaper corruption on resume from suspend with Nvidia GPUs
# reported at https://gitlab.gnome.org/GNOME/gnome-shell/issues/1084
# and due to this feature of the Nvidia driver:
# https://www.khronos.org/registry/OpenGL/extensions/NV/NV_robustness_video_memory_purge.txt
# Adapted from Paul Knopf (https://gitlab.gnome.org/GNOME/gnome-shell/issues/1084)
# and via https://bugs.launchpad.net/ubuntu/+source/mutter/+bug/1809407/comments/22
# Note that this does show the user's screen when coming out of sleep briefly.
# machinectl can be found in the systemd-container package on Ubuntu.

PROGNAME=$(/usr/bin/basename "$0")
state=$1
action=$2

function log {
    /usr/bin/logger -i -t "$PROGNAME" "$*"
}

log "Running $action $state"
if [[ $state == post ]]; then
    log "WAKE UP"
    pid_Gnome=($(/usr/bin/pgrep gnome-session))
    pid_Gnome=("${pid_Gnome[@]%%:*}")
    command="/usr/bin/dbus-send --type=method_call --dest=org.gnome.Shell /org/gnome/Shell org.gnome.Shell.Eval "string:global.reexec_self\(\)""
        for i in "${pid_Gnome[@]}"
    do
        user=$(/bin/ps -o user= -p $i)
        log $i $user
        if [ "$user" != "gdm" ]; then
            log "machinectl shell $user@.host $command"
            /bin/machinectl shell $user@.host $command
        fi
    done
fi
exit 0

Revision history for this message
Daniel van Vugt (vanvugt) wrote :

The formal fix is awaiting review;

  https://gitlab.gnome.org/GNOME/mutter/merge_requests/600

but you can download it as a patch:

  https://gitlab.gnome.org/GNOME/mutter/merge_requests/600.patch

Revision history for this message
slimbook (slimbook) wrote :

Hello,
A few weeks ago we fixed this bug with this script:

https://slimbook.es/tutoriales/linux/422-fixing-the-broken-desktop-after-resuming-from-a-suspension-in-gnome

I hope it helps someone.

Revision history for this message
Daniel van Vugt (vanvugt) wrote :

The official fix landed upstream last night. It is fixed in mutter version 3.33.4 and later.

Revision history for this message
Francisco Lopes (oblita) wrote :

@vanvugt can this fix be backported to gnome-3-32 branch? I just talked to the package maintainer in Arch repos and they said 3.33.4 is odd numbered, hence not a stable release, so we would have to wait for 3.34, but given how this is affecting many people I wished it sooner rather than later. They said they track 3.32, so if it gets backported, the package registry will pick it.

Changed in gnome-shell:
importance: Unknown → Undecided
importance: Undecided → Unknown
status: New → Unknown
Revision history for this message
Daniel van Vugt (vanvugt) wrote :

I want to give the fix time to settle, because the number of sub-fixes required seemed to change between the 3.33 source and 3.32. And aside from that, we are required to get the fix into 19.10 first.

So I will probably wait for this fix to enter 19.10 via mutter 3.34.0 (when it is ready). And then will assess backports to 3.32.

tags: removed: cosmic
Revision history for this message
WinEunuchs2Unix (ricklee518) wrote :

I have three screens:

1) nVidia GTX 970M HDMI port of laptop. 1920x1080
2) USB-C, Thunderbolt III HDMI dongle. 3840x2160
3) Laptop eDP1 (i7 6700HQ, HD 530 iGPU). 1920x1080

Upon resume the 3840x2160 screen's wallpaper is **OK**.

The two 1920x1080 screens wallpaper is all white.

Ubuntu 16.04.6 LTS upgrade direct to 19.04 w/kernel 5.0.0-23 yesterday.

Revision history for this message
Diego (diego-giglio) wrote :

Sorry for inconvenient Daniel, but the version 3.33.4 was rejected in review?

Revision history for this message
Daniel van Vugt (vanvugt) wrote :

The fix for this bug is in mutter version 3.33.4, which will eventually reach Ubuntu version 19.10, and then maybe others after that.

Nothing has been rejected.

Revision history for this message
Launchpad Janitor (janitor) wrote :
Download full text (4.5 KiB)

This bug was fixed in the package mutter - 3.33.91-1ubuntu1

---------------
mutter (3.33.91-1ubuntu1) eoan; urgency=medium

  [ Iain Lane ]
  * Merge with debian. Remaining changes:
    + debian/control:
      - Update VCS flags to point to launchpad
      - Update maintainer to ubuntu
    + debian/gbp.conf: update branch to point to ubuntu/master
    + debian/patches/x11-Add-support-for-fractional-scaling-using-Randr.patch:
      - X11: Add support for fractional scaling using Randr

  [ Marco Trevisan (Treviño) ]
  * debian/patches: Refresh. Compositor is now a GObject, so we can use
    smarter connection

mutter (3.33.91-1) experimental; urgency=medium

  [ Dmitry Shachnev ]
  * Update dh_girepository argument from mutter-4 to mutter-5

  [ Iain Lane ]
  * New upstream release
    + Fix primary selection copy and paste between X11 and wayland
    + Improve monitor hotplug support
    + Remove a source of frame skips
    + Fix windows being lowered after unmaximizing with double click
    + Remove Clutter API for global grabs
    + Improve processing of incompressible events
    + Add xdg-output v3 support
  * control: Bump BDs per upstream
  * debian/libmutter-5-0.symbols: Update for new/dropped symbols.
    The ABI of libmutter-5-0 is *not* considered stable until the final
    release.

mutter (3.33.90-2) experimental; urgency=medium

  [ Marco Trevisan (Treviño) ]
  * debian/rules: Use more verbose logging, printing failure output

  [ Iain Lane ]
  * tests-Tag-closed-transient-no-input-tests-as-flaky.patch, rules: Tag some
    tests as 'flaky' and run them non-fatally.
  * rules: Add mips64el to mips arches to not run tests on

mutter (3.33.90-1) experimental; urgency=medium

  [ Marco Trevisan (Treviño) ]
  * New upstream release
    + Avoid repainting covered areas
    + Start Xwayland on demand
    + Expose layout manager properties to transitions
  * debian/patches: Refresh

  [ Simon McVittie ]
  * d/libmutter-5-0.bug-control: Include GL drivers in bug reports

mutter (3.33.4-1) experimental; urgency=medium

  * New upstream release
    + Add API to reorder workspaces
    + Add a Sysprof-based profiler
    + Add initial KMS transactional support
    + Add xdg-output v2 support
    + Consolidate frame throttling
    + Defer actor allocation till shown
    + Discard page flip retries on hotplug
    + Don't emit ::size-changed when only position changed
    + Don't use grab modifiers when shortcuts are inhibited
    + Expose workspace layout properties
    + Fix background corruption on Nvidia after resuming from suspend (LP:
      #1809407)
    + Fix black shadows when using fractional scaling
    + Fix modifier-drag on wayland subsurfaces
    + Fix running X11 applications with sudo under wayland
    + Fix setting blank cursor under wayland
    + Fix stuttering due to unchanged power save mode notifications
    + Fix text selection color rendering
    + Handle returning from fullscreen/maximization better
    + Honor startup sequence workspace on wayland
    + Implement locate-pointer accessibility feature
    + Implement mouse accessibility
    + Implement toggle-keys notification
    + Improve finding new focus window when the ...

Read more...

Changed in mutter (Ubuntu Eoan):
status: In Progress → Fix Released
Revision history for this message
Rafael (marinho-eo) wrote :
Revision history for this message
Daniel van Vugt (vanvugt) wrote :

Ubuntu 19.10: Just update your system.
Ubuntu 19.04: Please wait for future updates.
Ubuntu 18.04: Please wait for future updates.

Changed in gnome-shell:
status: Unknown → Fix Released
Changed in mutter (Ubuntu Disco):
status: Confirmed → In Progress
Revision history for this message
satmandu (satadru-umich) wrote :

I'm still seeing this issue on resume from sleep today on my updated 19.10 setup with gnome-shell 3.34.0-1ubuntu1.

For what it is worth I also see this in my dmesg at boot:

[ 19.404177] gnome-shell[3485]: segfault at 18 ip 00007f82407d4924 sp 00007ffeb4e50f08 error 4 in libmutter-5.so.0.0.0[7f82406e7000+f2000]
[ 19.404192] Code: c3 0f 1f 44 00 00 f3 0f 1e fa 8b 47 30 c3 0f 1f 84 00 00 00 00 00 f3 0f 1e fa 8b 47 34 c3 0f 1f 84 00 00 00 00 00 f3 0f 1e fa <48> 8b 47 18 c3 0f 1f 80 00 00 00 00 f3 0f 1e fa 41 54 53 48 89 fb

Revision history for this message
Daniel van Vugt (vanvugt) wrote :

satmandu,

Please report that crash as a new bug using these instructions:
https://wiki.ubuntu.com/Bugs/Responses#Missing_a_crash_report_or_having_a_.crash_attachment

description: updated
description: updated
description: updated
description: updated
description: updated
description: updated
Revision history for this message
Daniel van Vugt (vanvugt) wrote :

Progress!

The fix has been accepted upstream in the 3.32 maintenance branch and a fix for Ubuntu 19.04 has also been proposed now: https://code.launchpad.net/~vanvugt/ubuntu/+source/mutter/+git/mutter/+merge/372984

Revision history for this message
Mike Kasberg (mkasberg) wrote :

Glad to see this proposed as a fix for 19.04, thanks for the work!

Changed in mutter (Ubuntu Bionic):
status: Confirmed → New
status: New → Incomplete
Changed in mutter (Ubuntu Bionic):
status: Incomplete → Confirmed
Revision history for this message
Daniel van Vugt (vanvugt) wrote :

Curious. I was looking at preparing a fix for Ubuntu 18.04 but in my initial setup for backporting have found that I can't reproduce the bug with Gnome 3.28. Though this is now with Nvidia driver 430.50 too.

Is anyone seeing this bug in Ubuntu 18.04? If so then please tell me what Nvidia driver you are using.

Changed in mutter (Ubuntu Bionic):
status: Confirmed → Incomplete
Revision history for this message
Daniel van Vugt (vanvugt) wrote :

Set up a new bionic machine and tested both Nvidia 430 and 340. The bug doesn't exist there.

Changed in mutter (Ubuntu Bionic):
status: Incomplete → Invalid
Revision history for this message
Diego (diego-giglio) wrote :

Hello, Daniel. I've experienced it in 18.04 and 19.04 both with nvidia-390.129 in a XPS with hybrid intel x nvidia.
I hope had contributed. If you need some thing let me know.

Revision history for this message
Daniel van Vugt (vanvugt) wrote :

I've now tried nvidia-390 on 18.04 and still can't reproduce the bug there.

Diego, can you provide a screenshot or photo of the corruption you see in 18.04?

Revision history for this message
Daniel van Vugt (vanvugt) wrote :

There appear to be no bug reports from 18.04 either.

Although I could make a patch for 18.04, it would never get released unless someone was able to reproduce the bug and verify it.

Changed in mutter (Ubuntu Bionic):
status: Invalid → Incomplete
Revision history for this message
Shane Synan (digitalcircuit) wrote :

I've had this issue in the past on Ubuntu 18.04 with the NVidia 340.x driver series. Two days ago with driver 340.107 and the latest released system updates, the lockscreen wallpaper corrupted for about 2 seconds on resume from suspend, then the wallpaper reloaded, appearing fine. In the past, the lockscreen wallpaper would remain corrupted for at least a minute (presumably indefinitely, but I didn't wait to find out).

This issue might be somehow fixed, but I'll keep a look out to see if it happens again and stays corrupted for longer than a minute.

Revision history for this message
Daniel van Vugt (vanvugt) wrote :

Dropped the bionic task until such time as this becomes a reproducible and testable bug there.

no longer affects: mutter (Ubuntu Bionic)
Changed in mutter (Ubuntu Disco):
status: In Progress → Won't Fix
Revision history for this message
Stuardo -StR- Rodríguez (stuardo) wrote :

OK, this is how the screen looks with latest Ubuntu 19.10. It's still a problem.

Revision history for this message
rigues (rigues) wrote :

I am still experiencing this on 19.10, with *both* the Nvidia and Nouveau drivers (maybe a new issue?). I'm currently running Nouveau, and when resuming from sleep the screen is completely corrupt, the only distinguishable feature is the mouse cursor. I tried logging in "blindly", but I don't know if I succeeded, as the only thing I can see is color noise.

Running Ubuntu 19.10 with kernel 5.3.0-23-generic, on an Intel Core i5 M460, GeForce 310M, Nvidia Driver 340.107 from nvidia-drivers. If you need any more info, just let me know.

Revision history for this message
Shane Synan (digitalcircuit) wrote :

Having recently tried upgrading an 18.04 system to 19.10, I can confirm ~rigues 's behavior - complete screen corruption after resuming from suspend when using the Nouveau driver (NVidia driver only partly corrupts).

However, this (corrupted screen, no workaround) might be different enough from a corrupted wallpaper (which can be dismissed) that it warrants a separate/new bug report?

Revision history for this message
manojav (manojav) wrote :

Nvidia Version 340.107

No LSB modules are available.
Distributor ID: Ubuntu
Description: Pop!_OS 19.10
Release: 19.10
Codename: eoan

mutter 3.34.1+git20191107-1ubuntu1~19.10.1pop0~1573683812~19.10~50e928a

Linux ###### 5.3.0-22-generic #24+system76~1573659475~19.10~26b2022-Ubuntu SMP Wed Nov 13 20:0 x86_64 x86_64 x86_64 GNU/Linux

Revision history for this message
manojav (manojav) wrote :

Also workaround is to restart gnome, Alt+F2, "r" and it seems to refresh the wallpaper.

Revision history for this message
Daniel van Vugt (vanvugt) wrote :

This bug is closed, rightly or wrongly. So if you continue to experience any problems then you will need to open a new bug by running:

  ubuntu-bug mutter

Revision history for this message
Jose Bartolome (zubzou) wrote :

I only want to add that the bug still exists. Since I don't know if the patch is on the way I don't know if it is worth to add a new bug. I am using Ubuntu 19.10 with Nvidia drive 340.107. Attached find a hardcopy of the screen

Revision history for this message
Daniel van Vugt (vanvugt) wrote :

This bug is closed. Some people have said they still see an issue though, so please use bug 1855757 from now on...

Revision history for this message
Juan Carlos Carvajal Bermúdez (jucajuca) wrote :
Revision history for this message
mistr (mstrecke) wrote :

This bug is still present in a fresh install of Ubuntu 20.04.1 LTS on a Lenovo T510.

Nvidia driver is 340.108

The work-around mentioned in #73 is working.

Revision history for this message
Daniel van Vugt (vanvugt) wrote :

This bug is closed. Please use bug 1855757 instead.

Revision history for this message
Sujit Kumar (therealsujitk) wrote :

For now you can easily fix this without any third-party applications by following these instructions: https://askubuntu.com/a/1269431/1119550

Revision history for this message
Daniel van Vugt (vanvugt) wrote :

^^^
That is bug 1892440. It is not relevant here.

To post a comment you must log in.