Launchpad has imported 16 comments from the remote bug at https://bugzilla.gnome.org/show_bug.cgi?id=787568.
If you reply to an imported comment from within Launchpad, your comment will be sent to the remote bug automatically. Read more about Launchpad's inter-bugtracker facilities at https://help.launchpad.net/InterBugTracking. ------------------------------------------------------------------------ On 2017-09-12T01:16:46+00:00 Adamw-c wrote: There's some discussion of this in https://bugzilla.gnome.org/show_bug.cgi?id=787240 , but that's really about another crash, so I figure this deserves its own report. Julian Andres Klode (in #787240), myself and Mikhail Gavrilov (in https://bugzilla.redhat.com/show_bug.cgi?id=1490072 ), and Jean-Baptiste Lallement (in https://bugs.launchpad.net/ubuntu/+source/gnome- shell/+bug/1715330 ) have all seen gnome-shell crash this way (taking down our entire sessions, natch) - a segfault in mutter _cogl_boxed_value_set_x , apparently because somehow it's getting called with NULL as its target. Myself, Julian and Mikhail saw it when opening / starting VMs in virt-manager and Boxes; the Ubuntu reproducer seems to possibly involve some kind of Ubuntu-specific update thing, but these are the cited steps: Test Case: Pre-requisites: Package update that downloads a payload and download fails (ie flash plugin) 1. Wait until the update-notifier dialog informing the user about the download failure shows up 2. Click on the 'Execute' buitton 3. Enter your credentials 4. Proceed with the download Expected result The package is downloaded Actual result This crash happens when the authentication window is displayed Myself, Mikhail and Jean-Baptiste all reported from 3.25.91, but I think Julian may have tested with something more recent. I don't see any changes between 3.25.91 and 3.25.92 which really look like they ought to fix this in any case. Reply at: https://bugs.launchpad.net/ubuntu/+source/mutter/+bug/1715330/comments/8 ------------------------------------------------------------------------ On 2017-09-12T04:13:13+00:00 Jonas Ådahl wrote: The bug is in gnome-shell, so moving there. Reply at: https://bugs.launchpad.net/ubuntu/+source/mutter/+bug/1715330/comments/10 ------------------------------------------------------------------------ On 2017-09-12T04:25:03+00:00 Jonas Ådahl wrote: Created attachment 359569 inhibitShortcuts: Don't destroy actor when hiding The same dialog can be shown multiple times; to not crash when that happens we need to avoid destroying the actor when hiding, otherwise we'll crash when running code that assumes it is still valid. Reply at: https://bugs.launchpad.net/ubuntu/+source/mutter/+bug/1715330/comments/11 ------------------------------------------------------------------------ On 2017-09-12T04:27:26+00:00 Jonas Ådahl wrote: To actually reproduce this reliably, the patch from bug 787570 is needed. On my set up, this exposes a gtk+/virt-manager bug resulting in the inhibit request being created and destroyed repeatedly. Reply at: https://bugs.launchpad.net/ubuntu/+source/mutter/+bug/1715330/comments/12 ------------------------------------------------------------------------ On 2017-09-12T04:28:08+00:00 Jonas Ådahl wrote: Uh, must have slipped. Reopening. Reply at: https://bugs.launchpad.net/ubuntu/+source/mutter/+bug/1715330/comments/13 ------------------------------------------------------------------------ On 2017-09-12T07:04:47+00:00 Ofourdan wrote: The issue is that virt-manager issues a grab() when it gains keyboard focus, and an ungrab() when it loses the keyboard focus. On Wayland, the grab()/ungrab() being wired to the shortcut inhibit mechanism, which in turn shows the dialog, therefore takes focus away from the virt-manager causes the continuous flickering. The solution, or at least a compromise, is to keep the dialog around even if the shortcut inhibit is cancelled by the client, so that the user is forced to make a choice that we can reuse on the next request. Reply at: https://bugs.launchpad.net/ubuntu/+source/mutter/+bug/1715330/comments/14 ------------------------------------------------------------------------ On 2017-09-12T08:33:34+00:00 Ofourdan wrote: Created attachment 359585 [PATCH] wayland: Keep the inhibit shortcut dialog On Wayland, the grab()/ungrab() in gtk+/gdk are wired to the shortcut inhibitor mechanism, which in turn shows the dialog, which can take focus away from the client window when the dialog is shown. If the client issues an ungrab() when the keyboard focus is lost, we would hide the dialog, causing the keyboard focus to be returned to the client surface, which in turn would issue a new grab(), so forth and so on, causing a continuous show/hide of the shortcut inhibitor dialog. To avoid this issue, keep the dialog around even if the shortcut inhibit is canceled by the client, so that the user is forced to make a choice that we can reuse on the next request without showing the dialog again. Instead of hiding the dialog when the shortcut inhibitor is destroyed by the client, we simply mark the request as canceled and do not apply the user's choice. Reply at: https://bugs.launchpad.net/ubuntu/+source/mutter/+bug/1715330/comments/15 ------------------------------------------------------------------------ On 2017-09-12T08:34:14+00:00 Ofourdan wrote: Created attachment 359586 [PATCH] wayland: do not leak shortcut inhibit data We would free the shortcut inhibit data only when the client destroys its request, which is not the case when the clients itself is destroyed, leading to a leak of the shortcut inhibit data. Free the data on resource destruction instead, and simply destroy the resource on destroy request. Reply at: https://bugs.launchpad.net/ubuntu/+source/mutter/+bug/1715330/comments/16 ------------------------------------------------------------------------ On 2017-09-12T08:49:34+00:00 Jonas Ådahl wrote: Review of attachment 359585: Looks good; just one nit about readability. ::: src/wayland/meta-wayland-inhibit-shortcuts-dialog.c @@ +85,3 @@ + if (data->request_canceled) + return; + Would make more sense if you put this condition in the dialog-response function. This being processed from the auto-approve path makes no sense. Reply at: https://bugs.launchpad.net/ubuntu/+source/mutter/+bug/1715330/comments/17 ------------------------------------------------------------------------ On 2017-09-12T08:50:15+00:00 Jonas Ådahl wrote: Review of attachment 359586: LGTM. Reply at: https://bugs.launchpad.net/ubuntu/+source/mutter/+bug/1715330/comments/18 ------------------------------------------------------------------------ On 2017-09-12T09:48:34+00:00 Ofourdan wrote: Created attachment 359592 [PATCH v2] wayland: Keep the inhibit shortcut dialog (In reply to Jonas Ådahl from comment #8) > Review of attachment 359585 [details] [review]: > > Looks good; just one nit about readability. > > ::: src/wayland/meta-wayland-inhibit-shortcuts-dialog.c > @@ +85,3 @@ > + if (data->request_canceled) > + return; > + > > Would make more sense if you put this condition in the dialog-response > function. This being processed from the auto-approve path makes no sense. Yeap, that also simplifies the patch... Reply at: https://bugs.launchpad.net/ubuntu/+source/mutter/+bug/1715330/comments/19 ------------------------------------------------------------------------ On 2017-09-12T13:04:27+00:00 Jonas Ådahl wrote: Review of attachment 359592: lgtm. Reply at: https://bugs.launchpad.net/ubuntu/+source/mutter/+bug/1715330/comments/20 ------------------------------------------------------------------------ On 2017-09-14T07:36:51+00:00 Ofourdan wrote: Comment on attachment 359586 [PATCH] wayland: do not leak shortcut inhibit data attachment 359586 pushed to git master as commit 2bf7974 - wayland: do not leak shortcut inhibit data Reply at: https://bugs.launchpad.net/ubuntu/+source/mutter/+bug/1715330/comments/21 ------------------------------------------------------------------------ On 2017-09-14T07:37:35+00:00 Ofourdan wrote: Comment on attachment 359592 [PATCH v2] wayland: Keep the inhibit shortcut dialog attachment 359592 pushed to git master as commit 9c16e4e - wayland: Keep the inhibit shortcut dialog Reply at: https://bugs.launchpad.net/ubuntu/+source/mutter/+bug/1715330/comments/22 ------------------------------------------------------------------------ On 2017-10-04T14:26:18+00:00 Ofourdan wrote: *** Bug 768959 has been marked as a duplicate of this bug. *** Reply at: https://bugs.launchpad.net/ubuntu/+source/mutter/+bug/1715330/comments/23 ------------------------------------------------------------------------ On 2017-11-03T02:48:28+00:00 Daniel van Vugt wrote: Those commit NNNN links aren't working. Can someone assign this bug to "mutter" to fix that? Reply at: https://bugs.launchpad.net/ubuntu/+source/mutter/+bug/1715330/comments/27 ** Changed in: mutter Status: Unknown => Fix Released ** Changed in: mutter Importance: Unknown => Critical ** Bug watch added: bugzilla.gnome.org/ #787240 https://bugzilla.gnome.org/show_bug.cgi?id=787240 -- You received this bug notification because you are a member of Desktop Packages, which is subscribed to mutter in Ubuntu. https://bugs.launchpad.net/bugs/1715330 Title: gnome-shell crashed with SIGSEGV in _cogl_boxed_value_set_x() Status in Mutter: Fix Released Status in mutter package in Ubuntu: Fix Released Bug description: https://errors.ubuntu.com/problem/923e1f8ff89aa3ff451c6aec260ec590152cf01a --- Artful Desktop daily, wayland session Test Case: Pre-requisites: Package update that downloads a payload and download fails (ie flash plugin) 1. Wait until the update-notifier dialog informing the user about the download failure shows up 2. Click on the 'Execute' buitton 3. Enter your credentials 4. Proceed with the download Expected result The package is downloaded Actual result This crash happens when the authentication window is displayed ProblemType: Crash DistroRelease: Ubuntu 17.10 Package: gnome-shell 3.25.91-0ubuntu3 ProcVersionSignature: Ubuntu 4.12.0-12.13-generic 4.12.8 Uname: Linux 4.12.0-12-generic x86_64 ApportVersion: 2.20.7-0ubuntu1 Architecture: amd64 CurrentDesktop: ubuntu:GNOME Date: Wed Sep 6 09:41:48 2017 DisplayManager: gdm3 ExecutablePath: /usr/bin/gnome-shell GsettingsChanges: b'org.gnome.shell' b'had-bluetooth-devices-setup' b'true' b'org.gnome.shell' b'favorite-apps' b"['org.gnome.Nautilus.desktop', 'firefox.desktop', 'google-chrome-beta.desktop', 'streamtuner2.desktop']" b'org.gnome.desktop.interface' b'gtk-im-module' b"'gtk-im-context-simple'" InstallationDate: Installed on 2014-07-23 (1140 days ago) InstallationMedia: Ubuntu 14.04.1 LTS "Trusty Tahr" - Release amd64 (20140722.2) ProcCmdline: /usr/bin/gnome-shell ProcEnviron: PATH=(custom, no user) XDG_RUNTIME_DIR=<set> LANG=fr_FR.UTF-8 SHELL=/bin/bash SegvAnalysis: Segfault happened at: 0x7f72063804ef: mov 0x8(%rdi),%eax PC (0x7f72063804ef) ok source "0x8(%rdi)" (0x00000008) not located in a known VMA region (needed readable region)! destination "%eax" ok SegvReason: reading NULL VMA Signal: 11 SourcePackage: gnome-shell StacktraceTop: ?? () from /usr/lib/x86_64-linux-gnu/mutter/libmutter-cogl-1.so ?? () from /usr/lib/x86_64-linux-gnu/mutter/libmutter-cogl-1.so ffi_call_unix64 () from /usr/lib/x86_64-linux-gnu/libffi.so.6 ffi_call () from /usr/lib/x86_64-linux-gnu/libffi.so.6 ?? () from /usr/lib/libgjs.so.0 Title: gnome-shell crashed with SIGSEGV in ffi_call_unix64() UpgradeStatus: Upgraded to artful on 2017-06-13 (84 days ago) UserGroups: adm dialout libvirt lpadmin lxd sambashare sudo To manage notifications about this bug go to: https://bugs.launchpad.net/mutter/+bug/1715330/+subscriptions -- Mailing list: https://launchpad.net/~desktop-packages Post to : [email protected] Unsubscribe : https://launchpad.net/~desktop-packages More help : https://help.launchpad.net/ListHelp
