Description of problem: When switching to a tty and back, gnome-shell immediately crashes. Steps to reproduce: 1. log in into a gnome+wayland session (gnome+x11 untested) 2. switch to any tty, e.g. tty3. It does not matter whether you login or not. 3. switch back to the tty with gnome session (usually tty2) What happens: Gnome-shell immediately crashes What should happen: No crash Version-Release number of selected component: gnome-shell-3.22.0-1.fc25 Additional info: reporter: libreport-2.8.0 backtrace_rating: 4 cmdline: /usr/bin/gnome-shell crash_function: wl_resource_post_event executable: /usr/bin/gnome-shell kernel: 4.8.0-0.rc7.git0.1.fc25.x86_64 pkg_fingerprint: 4089 D8F2 FDB1 9C98 pkg_vendor: Fedora Project runlevel: N 5 type: CCpp Truncated backtrace: Thread no. 1 (10 frames) #0 wl_resource_post_event at src/wayland-server.c:194 #1 wl_pointer_send_motion at /usr/include/wayland-server-protocol.h:3423 #2 meta_wayland_pointer_send_motion at wayland/meta-wayland-pointer.c:339 #3 notify_motion at wayland/meta-wayland-pointer.c:557 #4 handle_motion_event at wayland/meta-wayland-pointer.c:564 #5 meta_wayland_pointer_handle_event at wayland/meta-wayland-pointer.c:710 #6 meta_wayland_seat_handle_event at wayland/meta-wayland-seat.c:360 #7 meta_wayland_compositor_handle_event at wayland/meta-wayland.c:208 #8 meta_display_handle_event at core/events.c:386 #9 event_callback at core/events.c:401
Created attachment 1204453 [details] File: backtrace
Created attachment 1204454 [details] File: core_backtrace
Created attachment 1204455 [details] File: dso_list
Created attachment 1204456 [details] File: limits
Created attachment 1204457 [details] File: namespaces
Created attachment 1204458 [details] File: proc_pid_status
This bug is 100% reproducible after installing the Gnome 3.22 megaupdate https://bodhi.fedoraproject.org/updates/FEDORA-2016-77ffa78abb
(In reply to Christian Stadelmann from comment #7) > This bug is 100% reproducible after installing the Gnome 3.22 megaupdate > https://bodhi.fedoraproject.org/updates/FEDORA-2016-77ffa78abb To be more precise: I think it is a regression introduced by the 3.22 megaupdate. Downgrading gnome-shell to 3.21.92-1.fc25.x86_64 makes the problem go away. This issue is 100% reproducible if animations are enabled. Workaround: disable animations.
(In reply to Christian Stadelmann from comment #8) > Workaround: disable animations. In fact, that's not reliable. Still getting the crash with animations disabled. Anyway, it seems like this crash is triggered by moving the mouse after switching from tty to gnome/wayland session. If you don't move your mouse, no crash happens. This issue is specific to gnome/wayland sessions and does not happen on gnome/X11 sessions.
Despite updating to mutter-3.22.0-2.fc25, etc., crashes caused by vt switches are still happening which the bug reporting tool redirects to this bug 1379098.
(In reply to Joachim Frieben from comment #10) > Despite updating to mutter-3.22.0-2.fc25, etc., crashes caused by vt > switches are still happening which the bug reporting tool redirects to this > bug 1379098. Same issue here.
Yep, just saw this after switching to a VT and back, with all current F25 packages from updates-testing plus: gtk3-3.22.1-2.fc25.x86_64 gnome-shell-3.22.1-1.fc25.x86_64 mutter-3.22.1-1.fc25.x86_64 proposing as a Final blocker, per criterion "All known bugs that can cause corruption of user data must be fixed or documented at Common F25 bugs.", reading 'corruption' to include 'loss' (as it's kinda intended to, at least by me, who wrote it) - switching to a VT is something people commonly do quite frequently and 'casually' (no-one expects the desktop session to suddenly crash, so they usually don't save all unsaved work in the desktop session before switching), and this crash could lead to the loss of all unsaved data from the desktop session. Christian, could you check if there's really any secret stuff in your attachments and set the bug public if not? abrt's 'private data' check very often gives false positives (it just looks for key phrases like 'password' and 'key' and stuff and is easily fooled by the same characters cropping up in other contexts, like library filenames).
(In reply to Adam Williamson from comment #12) > proposing as a Final blocker, per criterion "All known bugs that can cause > corruption of user data must be fixed or documented at Common F25 bugs.", > reading 'corruption' to include 'loss' (as it's kinda intended to, at least > by me, who wrote it) - switching to a VT is something people commonly do > quite frequently and 'casually' (no-one expects the desktop session to > suddenly crash, so they usually don't save all unsaved work in the desktop > session before switching), and this crash could lead to the loss of all > unsaved data from the desktop session. +1 (if I'm allowed to vote). > Christian, could you check if there's really any secret stuff in your > attachments and set the bug public if not? abrt's 'private data' check very > often gives false positives (it just looks for key phrases like 'password' > and 'key' and stuff and is easily fooled by the same characters cropping up > in other contexts, like library filenames). Ok. Feel free to make this bug public, i.e. remove it from the "Fedora Contrib Private" group. I don't have permission to do so.
Similar problem has been detected: Again tried to reproduce bug 1379098, following the same steps as described in bug 1379440 comment 26. reporter: libreport-2.8.0 backtrace_rating: 4 cmdline: /usr/bin/gnome-shell crash_function: wl_resource_post_event executable: /usr/bin/gnome-shell global_pid: 17017 kernel: 4.8.1-1.fc25.x86_64 package: gnome-shell-3.22.1-1.fc25 pkg_fingerprint: 4089 D8F2 FDB1 9C98 pkg_vendor: Fedora Project reason: gnome-shell killed by SIGSEGV runlevel: N 5 type: CCpp uid: 1000
Discussed during the 2016-10-17 blocker review meeting: [1] The decision to classify this bug as an "AcceptedBlocker" was made as it violates the following criteria: "All known bugs that can cause corruption of user data must be fixed or documented at Common F25 bugs". [1] https://meetbot.fedoraproject.org/fedora-blocker-review/2016-10-17/f25-blocker-review.2016-10-17-16.02.txt
I can easily reproduce this on a bare metal machine with Intel graphics. GNOME crashes immediately after VT switch to text and back to graphics and *moving the cursor*. On my laptop, it takes longer to reproduce this (I have to switch multiple times), but it also occurs. So it seems this is pretty universal (Adam reproduced it with nouveau). I don't seem to hit it in a VM, though.
I reproduced this as well, but it seems to only reproduce when an application is open on the desktop. I can't reproduce the behavior immediately after a session start, when no apps are open on the desktop.
Sorry, details on current packages: mutter-3.22.1-3.fc25.x86_64 gtk3-3.22.1-2.fc25.x86_64 gnome-shell-3.22.1-1.fc25.x86_64
There's an update for this issue here: https://bodhi.fedoraproject.org/updates/FEDORA-2016-1544cc87d7 FEDORA-2016-1544cc87d7 Submitter rtcm Builds mutter-3.22.1-5.fc25 Update ID FEDORA-2016-1544cc87d7 Release F25 Karma 0 stable threshold: 3 unstable threshold: -3 Type bugfix Severity unspecified Request testing Status pending Suggestion unspecified Submitted 21 hours ago, 2016-10-20 17:19:45 (UTC)
mutter-3.22.1-5.fc25 has been pushed to the Fedora 25 testing repository. If problems still persist, please make note of it in this bug report. See https://fedoraproject.org/wiki/QA:Updates_Testing for instructions on how to install test updates. You can provide feedback for this update here: https://bodhi.fedoraproject.org/updates/FEDORA-2016-1544cc87d7
mutter-3.22.1-5.fc25 has been pushed to the Fedora 25 stable repository. If problems still persist, please make note of it in this bug report.
(In reply to Fedora Update System from comment #21) > mutter-3.22.1-5.fc25 has been pushed to the Fedora 25 stable repository. Tested, seems fixed.