[Freeciv-Dev] [bug #24010] Weird display glitch: despotism heads on main map (after tileset change)

2015-12-12 Thread Jacob Nevins 
Follow-up Comment #10, bug #24010 (project freeciv):

For reference: I think it's probably this bug that manifests in the Windows
S2_5 r30387 test build

(which doesn't contain this fix) as a reliable crash after leaving and
restarting a client-spawned server, with a backtrace like this:


63A65B12  libgobject-2.0-0.dll:63A65B12  g_type_check_instance
63A65BEE  libgobject-2.0-0.dll:63A65BEE  g_type_check_is_value_type
63A685CA  libgobject-2.0-0.dll:63A685CA  g_value_type_compatible
63A5043A  libgobject-2.0-0.dll:63A5043A  g_object_steal_data
011EFC53  libgtk-win32-2.0-0.dll:011EFC53  gtk_link_button_get_visited
011F054D  libgtk-win32-2.0-0.dll:011F054D  gtk_list_store_set_valist
011F05BB  libgtk-win32-2.0-0.dll:011F05BB  gtk_list_store_set
004467AD  _editgui_tileset_changed
0040859C  _set_client_state
0046A9DB  _handle_start_phase
0047EA31  _client_handle_packet
0040701B  _client_packet_input
0042E4CB  _input_from_server
00402709  _get_net_input
68642DDC  libglib-2.0-0.dll:68642DDC  g_thread_yield
685F9630  libglib-2.0-0.dll:685F9630  g_source_is_destroyed
685FA214  libglib-2.0-0.dll:685FA214  g_main_context_dispatch
685FA3EF  libglib-2.0-0.dll:685FA3EF  g_main_context_dispatch
685FA7DE  libglib-2.0-0.dll:685FA7DE  g_main_loop_run
011F2C10  libgtk-win32-2.0-0.dll:011F2C10  gtk_main
00405F0C  _ui_main 
00407672  _client_main
0040235B  _console_main
00402418  _WinMain@16
00401AD6  _main
004010DB  ___mingw_CRTStartup
00401158  _WinMainCRTStartup
757DEE6C  C:\Windows\system32\kernel32.dll:757DEE6C  BaseThreadInitThunk
773D3AB3  C:\Windows\SYSTEM32\ntdll.dll:773D3AB3  RtlInitializeExceptionChain
773D3A86  C:\Windows\SYSTEM32\ntdll.dll:773D3A86  RtlInitializeExceptionChain


___

Reply to this item at:

  

___
  Message sent via/by Gna!
  http://gna.org/


___
Freeciv-dev mailing list
Freeciv-dev@gna.org
https://mail.gna.org/listinfo/freeciv-dev

[Freeciv-Dev] [bug #24010] Weird display glitch: despotism heads on main map (after tileset change)

2015-11-07 Thread Jacob Nevins 
Follow-up Comment #8, bug #24010 (project freeciv):

>> in practice: idle_callback doing things on inconsistent data
> To test that hypothesis: Does the attached patch file #2]
> (not a commit candidate) improve things?
Not obviously; I still had the same incorrect graphics drawn before and after
applying the patch to S2_5 r30455. (I didn't see console messages in this run
in either case)

> I think the double unref fixed by attached patch [file #25617]
> was behind all this.
This, however, does make my current symptom go away.

___

Reply to this item at:

  

___
  Message sent via/by Gna!
  http://gna.org/


___
Freeciv-dev mailing list
Freeciv-dev@gna.org
https://mail.gna.org/listinfo/freeciv-dev

[Freeciv-Dev] [bug #24010] Weird display glitch: despotism heads on main map (after tileset change)

2015-11-07 Thread Marko Lindqvist 
Update of bug #24010 (project freeciv):

  Status:  Ready For Test => Fixed  
 Assigned to:None => cazfi  
 Open/Closed:Open => Closed 


___

Reply to this item at:

  

___
  Message sent via/by Gna!
  http://gna.org/


___
Freeciv-dev mailing list
Freeciv-dev@gna.org
https://mail.gna.org/listinfo/freeciv-dev

[Freeciv-Dev] [bug #24010] Weird display glitch: despotism heads on main map (after tileset change)

2015-11-04 Thread Marko Lindqvist 
Update of bug #24010 (project freeciv):

  Status:None => Ready For Test 
 Planned Release: 2.5.2, 2.6.0, 3.0.0 => 2.5.2  

___

Follow-up Comment #7:

I think the double unref fixed by attached patch was behind all this.

Only S2_5 was affected by this. This should have been part of bug #23935, but
when backporting the fix to S2_5 the fact that there was one case more to fix
than in later branches went unnoticed.

(file #25617)
___

Additional Item Attachment:

File name: DoubleSpecialUnref.patch   Size:0 KB


___

Reply to this item at:

  

___
  Message sent via/by Gna!
  http://gna.org/


___
Freeciv-dev mailing list
Freeciv-dev@gna.org
https://mail.gna.org/listinfo/freeciv-dev

[Freeciv-Dev] [bug #24010] Weird display glitch: despotism heads on main map (after tileset change)

2015-11-01 Thread Jacob Nevins 
Update of bug #24010 (project freeciv):

 Planned Release: => 2.5.2, 2.6.0, 3.0.0
 Summary: Weird display glitch: despotism heads on main map
(after tileset change?) => Weird display glitch: despotism heads on main map
(after tileset change)

___

Follow-up Comment #4:

I've got similar symptoms to show up by following the same sorts of steps
again:
0 freeciv-gtk2 -t isotrident
0 start game
0 set tileset to amplio2

Editor and use of civ2civ3 don't seem to be necessary; symptoms seem to start
immediately after changing tileset.

Since it's reproducible, setting 2.5.2 target.

___

Reply to this item at:

  

___
  Message sent via/by Gna!
  http://gna.org/


___
Freeciv-dev mailing list
Freeciv-dev@gna.org
https://mail.gna.org/listinfo/freeciv-dev

[Freeciv-Dev] [bug #24010] Weird display glitch: despotism heads on main map (after tileset change)

2015-11-01 Thread Marko Lindqvist 
Follow-up Comment #5, bug #24010 (project freeciv):

I can't reproduce your exact problem, but:
1) For a long time now I've been getting client crashes when switching
tilesets multiple times
2) I do get "g_object_unref: assertion 'G_IS_OBJECT (object)' failed"
-warnings (maybe related to some of the recent g_object_unref() / free()
additions - so there might be a double g_object_unref() now)
3) I do *not* get the crashes when I run the client in debugger - so it might
be similar timing related issue we had with research dialog update (in
practice: idle_callback doing things on inconsistent data)

___

Reply to this item at:

  

___
  Message sent via/by Gna!
  http://gna.org/


___
Freeciv-dev mailing list
Freeciv-dev@gna.org
https://mail.gna.org/listinfo/freeciv-dev

[Freeciv-Dev] [bug #24010] Weird display glitch: despotism heads on main map (after tileset change)

2015-11-01 Thread Marko Lindqvist 
Follow-up Comment #6, bug #24010 (project freeciv):

> in practice: idle_callback doing things on inconsistent data

To test that hypothesis: Does the attached patch (not a commit candidate)
improve things?

(file #2)
___

Additional Item Attachment:

File name: UpdatesFrozenDuringTSChange.patch Size:1 KB


___

Reply to this item at:

  

___
  Message sent via/by Gna!
  http://gna.org/


___
Freeciv-dev mailing list
Freeciv-dev@gna.org
https://mail.gna.org/listinfo/freeciv-dev

[Freeciv-Dev] [bug #24010] Weird display glitch: despotism heads on main map (after tileset change?)

2015-11-01 Thread Jacob Nevins 
Update of bug #24010 (project freeciv):

 Summary: Weird display glitch: despotism heads on main map =>
Weird display glitch: despotism heads on main map (after tileset change?)

___

Follow-up Comment #3:

My console is full of


(freeciv-gtk2:31741): Gdk-CRITICAL **: IA__gdk_draw_pixbuf: assertion
'GDK_IS_PIXBUF (pixbuf)' failed


After quitting and restarting, the client finally segfaulted. (I have no
coredump.)

Probably relevant: tileset was isotrident in my saved settings; I had changed
to amplio2 through the client UI during this session (after the game started).

___

Reply to this item at:

  

___
  Message sent via/by Gna!
  http://gna.org/


___
Freeciv-dev mailing list
Freeciv-dev@gna.org
https://mail.gna.org/listinfo/freeciv-dev