Bug#861645: gsequencer FTBFS on mips/mipsel: FAIL: ags_xorg_application_context_test

[Joël Krähemann]

Hi again

Sorry wrong context. It is the problem for sure:
Since it is used by AgsAutomationEditor, too.

Joël


On Fri, May 5, 2017 at 6:31 PM, Joël Krähemann <jkraehemann at gmail.com> wrote:
> Hi
>
> The flag was set. However it is not that beautiful.
> So it can't be the problem.
>
> https://anonscm.debian.org/git/pkg-multimedia/gsequencer.git/tree/ags/X/ags_editor.c#n222
>
> Bests,
> Joël
>
>
> On Fri, May 5, 2017 at 6:09 PM, James Cowgill <jcowgill at debian.org> wrote:
>> Hi,
>>
>> On 05/05/17 13:36, Joël Krähemann wrote:
>>> Hi
>>>
>>> finally I got a stack-trace
>>>
>>> Using host libthread_db library "/lib/mipsel-linux-gnu/libthread_db.so.1".
>>> Core was generated by
>>> `/home/jkraehemann/gsequencer-0.8.0/.libs/ags_xorg_application_context_test'.
>>> Program terminated with signal SIGSEGV, Segmentation fault.
>>> #0  0x75e2af68 in IA__gtk_widget_show_all (widget=0x7280b790) at
>>> ./gtk/gtkwidget.c:3376
>>> 3376 ./gtk/gtkwidget.c: No such file or directory.
>>> (gdb)
>>> (gdb) bt
>>> #0  0x75e2af68 in IA__gtk_widget_show_all (widget=0x7280b790) at
>>> ./gtk/gtkwidget.c:3376
>>> #1  0x7677b6c8 in ags_machine_selector_popup_new
>>> (machine_selector=0x728104c0) at
>>> ags/X/editor/ags_machine_selector.c:569
>>> #2  0x766b60a4 in ags_automation_editor_init
>>> (automation_editor=0x7276f238) at ags/X/ags_automation_editor.c:222
>>> #3  0x75660948 in g_type_create_instance () from
>>> /usr/lib/mipsel-linux-gnu/libgobject-2.0.so.0
>>> Backtrace stopped: frame did not save the PC
>>
>> I ran this test in gdb on a real mips machine. When tracing through
>> ags_machine_selector_popup_new it seems that the "keys" variable is
>> never initialized and a garbage pointer is passed to gtk_widget_show_all.
>>
>> The variable is never initialized because this condition is false:
>> (AGS_MACHINE_SELECTOR_SHOW_SHIFT_PIANO & (machine_selector->flags)) != 0
>>
>> Thanks,
>> James
>>