Gnome Shell crash on login with wayland
Hi,
I am trying run gnome @ wayland under Gentoo for hidpi scalling support on my dual-screen setup.
Running on latest available kernel (4.20.0), mesa (18.3.1) and mutter/clutter (3.26.2). Amdgpu stack. Since it's Gentoo, I am running at Gnome 3.26. This might be as well a bug in clutter itself, as it happens in clutter_stage_view_get_onscreen
, however, it gets invalid reference from mutter.
This particular crash (nullptr) seems to be not handled in current master, so maybe attached dump will be somehow helpful. I could upload complete coredump, but since its 212MB, I'd like to check whether this is a real issue and it's not fixed already.
coredump debug:
PID: 12513 (gnome-shell)
UID: 118 (gdm)
GID: 996 (gdm)
Signal: 11 (SEGV)
Timestamp: Sun 2018-12-30 18:38:47 CET (1min 3s ago)
Command Line: /usr/bin/gnome-shell
Executable: /usr/bin/gnome-shell
Control Group: /user.slice/user-118.slice/session-c16.scope
Unit: session-c16.scope
Slice: user-118.slice
Session: c16
Owner UID: 118 (gdm)
Boot ID: d59a706dbc4d4b19bb8294b0ed8d6397
Machine ID: 0e186f0aa240428bbde281b84d523263
Hostname: stacjonarka
Storage: /var/lib/systemd/coredump/core.gnome-shell.118.d59a706dbc4d4b19bb8294b0ed8d6397.12513.1546191527000000.lz4
Message: Process 12513 (gnome-shell) of user 118 dumped core.
GNU gdb (Gentoo 8.2.1 p1) 8.2.1
** CUT CLUTTER **
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib64/libthread_db.so.1".
Core was generated by `/usr/bin/gnome-shell'.
Program terminated with signal SIGSEGV, Segmentation fault.
#0 0x00007fe97a50fef7 in clutter_stage_view_get_onscreen (view=0x0) at clutter-stage-view.c:82
82 return priv->framebuffer;
[Current thread is 1 (Thread 0x7fe96aa73a00 (LWP 12513))]
(gdb) p *priv
Cannot access memory at address 0xffffffffffffffc0
(gdb) bt
#0 0x00007fe97a50fef7 in clutter_stage_view_get_onscreen (view=0x0) at clutter-stage-view.c:82
#1 0x00007fe97a186b78 in meta_renderer_native_queue_modes_reset (renderer_native=0x7fe954005df0) at backends/native/meta-renderer-native.c:1392
#2 0x00007fe97a187779 in meta_stage_native_rebuild_views (stage_native=stage_native@entry=0x562cc93c93b0) at backends/native/meta-renderer-native.h:34
#3 0x00007fe97a17d7f1 in meta_backend_native_update_screen_size (backend=<optimized out>, width=5760, height=2160) at backends/native/meta-backend-native.c:564
#4 0x00007fe97a0e5a83 in meta_backend_sync_screen_size (backend=backend@entry=0x562cc8d9f110) at backends/meta-backend-private.h:52
#5 0x00007fe97a0e623d in meta_backend_real_post_init (backend=0x562cc8d9f110) at backends/meta-backend.c:448
#6 0x00007fe97a17db2f in meta_backend_native_post_init (backend=0x562cc8d9f110) at backends/meta-backend-private.h:52
#7 0x00007fe97a0e675b in meta_backend_post_init (backend=<optimized out>) at backends/meta-backend-private.h:52
#8 meta_clutter_init () at backends/meta-backend.c:1036
#9 0x00007fe97a12dade in meta_init () at core/main.c:569
#10 0x0000562cc6fb859a in main (argc=<optimized out>, argv=<optimized out>) at ../gnome-shell-3.26.2/src/main.c:492
(gdb) p meta_renderer_native_queue_modes_reset::l
$1 = (GList *) 0x562cc9498d60
(gdb) p *meta_renderer_native_queue_modes_reset::l
$2 = {data = 0x0, next = 0x562cc9499200, prev = 0x0}
Wayland at KDE's is starting fine, so it should not be a hardware/config problem.