[oi-dev] lightdm failed repeatedly

[Marcel Telka]

On Sat, Nov 02, 2024 at 08:31:30PM -0500, Gary Mills wrote:
> <root at intel># mdb /mnt/core 
> Loading modules: [ libc.so.1 ld.so.1 ]
> > ::status
> debugging core file of lightdm (64-bit) from intel
> initial argv: /usr/sbin/lightdm
> threading model: native threads
> status: process terminated by SIGSEGV (Segmentation Fault), addr=c8
> > ::stack
> display_server_stopped_cb+0x2a9()
> libgobject-2.0.so.0.8200.2`_g_closure_invoke_va+0xd0()
> libgobject-2.0.so.0.8200.2`signal_emit_valist_unlocked+0x1173()
> libgobject-2.0.so.0.8200.2`g_signal_emit_valist+0x3d()
> libgobject-2.0.so.0.8200.2`g_signal_emit+0x7d()
> libgobject-2.0.so.0.8200.2`g_closure_invoke+0xc5()
> libgobject-2.0.so.0.8200.2`signal_emit_unlocked_R.isra.0+0xb86()
> libgobject-2.0.so.0.8200.2`signal_emit_valist_unlocked+0x10c1()
> libgobject-2.0.so.0.8200.2`g_signal_emit_valist+0x3d()
> libgobject-2.0.so.0.8200.2`g_signal_emit+0x7d()
> libglib-2.0.so.0.8200.2`g_child_watch_dispatch+0x66()
> libglib-2.0.so.0.8200.2`g_main_dispatch+0xb6()
> libglib-2.0.so.0.8200.2`g_main_context_iterate_unlocked.isra.0+0x2a8()
> libglib-2.0.so.0.8200.2`g_main_loop_run+0x117()
> main+0xcc7()
> _start_crt+0x87()
> _start+0x18()
> > ^d

The segfault happened here (apparently rax is zero):

> display_server_stopped_cb+0x2a9::dis
display_server_stopped_cb+0x27b:jne    -0x154 <display_server_stopped_cb+0x12d>
display_server_stopped_cb+0x281:nopl   0x0(%rax)
display_server_stopped_cb+0x288:movq   -0x38(%rbp),%rdi
display_server_stopped_cb+0x28c:movl   $0x41a2e8,%esi <PLT:g_object_unref>
display_server_stopped_cb+0x291:call   -0x1450e <PLT:g_list_free_full>
display_server_stopped_cb+0x296:movl   0x44(%r14),%edx
display_server_stopped_cb+0x29a:testl  %edx,%edx
display_server_stopped_cb+0x29c:jne    -0x1dd <display_server_stopped_cb+0xc5>
display_server_stopped_cb+0x2a2:movq   (%r12),%rax
display_server_stopped_cb+0x2a6:movq   %r12,%rdi
display_server_stopped_cb+0x2a9:call   *0xc8(%rax)    <------------- HERE
display_server_stopped_cb+0x2af:movq   %rax,%rdi
display_server_stopped_cb+0x2b2:testq  %rax,%rax
display_server_stopped_cb+0x2b5:je     +0xe <display_server_stopped_cb+0x2c5>
display_server_stopped_cb+0x2b7:call   +0x4724 <session_get_display_server>
display_server_stopped_cb+0x2bc:cmpq   %rax,%r13
display_server_stopped_cb+0x2bf:jne    -0x200 <display_server_stopped_cb+0xc5>
display_server_stopped_cb+0x2c5:call   -0x83da  <logger_get_type>
display_server_stopped_cb+0x2ca:movq   %r12,%rdi
display_server_stopped_cb+0x2cd:movq   %rax,%rsi
display_server_stopped_cb+0x2d0:call   -0x146bd <PLT:g_type_check_instance_cast>
>

So the next step would be to dig into the lightdm sources.  Anyway, our
lightdm is 1.30.0 released over 5 years ago.  Maybe as a first step it
would be better to update to 1.32.0 (released two years ago) and start
the debugging from there (if the core persists).

-- 
+-------------------------------------------+
| Marcel Telka   e-mail:   marcel at telka.sk  |
|                homepage: http://telka.sk/ |
+-------------------------------------------+