Segfault in a187.

[javajames]

[Ubuntu 12.04 with linux 3.2 generic 64bit and catalyst 12.6 drivers]

First installed saturday from the executable download, ran SUMlauncher today and updated it from a186 to a187, at this point it worked absolutely fine, and i could play any map. First crash, but i don't think is related, happened when i built a block which overlapped the player. I didn't think anything of it and it worked fine when i restarted. Then i started eroded plateau, and it worked fine, however i started to fight another character, when (abnormal) my frame rate dropped from 200 to 10 and stayed such for about 5 seconds, before the game crashed and quit (i don't have logs of that), on restarting every map i start the game crashes while loading, (later to be discovered as a segfault), i reinstalled it using the a186 installer and then ran the updater, the exact same thing happened. Openned up terminal and ran it with gdb, and this was the outcome:

Code: Select all

:/usr/local/games/Overgrowth$ gdb overgrowth.bin.x86_64
GNU gdb (Ubuntu/Linaro 7.4-2012.04-0ubuntu2) 7.4-2012.04
Copyright (C) 2012 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "x86_64-linux-gnu".
For bug reporting instructions, please see:
<http://bugs.launchpad.net/gdb-linaro/>...
Reading symbols from /usr/local/games/Overgrowth/overgrowth.bin.x86_64...(no debugging symbols found)...done.
(gdb) run
Starting program: /usr/local/games/Overgrowth/overgrowth.bin.x86_64 
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
[New Thread 0x7fffe76ed700 (LWP 13236)]
[Thread 0x7fffe76ed700 (LWP 13236) exited]
[New Thread 0x7fffe76ed700 (LWP 13237)]
[New Thread 0x7fffe2eeb700 (LWP 13238)]
RGBA bits, depth: 8 8 8 8, 24
Anti-aliasing samples: 2
32 texture units supported.
GL_MAX_VERTEX_UNIFORM_COMPONENTS: 16384
1 SDL joysticks found
HuiJia  USB Joystick
Hardware threads: 8
[New Thread 0x7fffdea5b700 (LWP 13240)]
Gtk-Message: Failed to load module "gtk-vector-screenshot"
[New Thread 0x7fffde25a700 (LWP 13241)]
[New Thread 0x7fffdcc72700 (LWP 13244)]
[New Thread 0x7fffdc471700 (LWP 13245)]
[New Thread 0x7fffdbc70700 (LWP 13246)]
[New Thread 0x7fffdb46f700 (LWP 13247)]
[New Thread 0x7fffdac6e700 (LWP 13248)]
[New Thread 0x7fffda46d700 (LWP 13249)]
[New Thread 0x7fffd9c6c700 (LWP 13250)]
[New Thread 0x7fffd8f7b700 (LWP 13264)]
[New Thread 0x7fffd1995700 (LWP 13268)]
*****************
Calculating detail textures
Navmesh: Adding mesh to sample...
Finished loading path set
"Loading" benchmarks:
59%: "Loading entities" (3 s)
3O%: "Loading terrain" (1 s)
O4%: "Loading navmesh" (26O ms)
O4%: "Loading shadows" (258 ms)
OO%: "Loading xml doc" (11 ms)
OO%: "Initializing" (151O us)
[Thread 0x7fffd1995700 (LWP 13268) exited]

Program received signal SIGSEGV, Segmentation fault.
0x0000000000af53d3 in asIScriptFunction::asIScriptFunction() ()

I have tried removing the entire /usr/local/games/Overgrowth folder, with the same outcome. I will also try to reinstall with the binary (nixinstaller) download from the forums.

This game was fun while it worked, any ideas?

[pure_krogan89]

i'm having the same problem and have found no answers. i suppose we'll just have to wait for the alpha that fixes it.

[Faw]

Seems to be a pretty common problem atm:

viewtopic.php?f=13&t=16075&p=171857#p171857

I would love to know what causes this.

[urkle]

FYI, David escalated this issue to me and I am hot on fixing it.. I actually believe I have it fixed this time.. It's "yet another fluke crash caused by Awesomium". Even though it crashes in angelscript.. it's caused by awesomium.. More specifically it's caused by chromium's use of an "alternate" malloc that seems to randomly be used or not used and causing crashes depending on some random even during linking.

Anyways.. disabling the alternate malloc so far has fixed the issue. I just have to go through the excruciating process of recompiling awesomium on 32bit now. (More time consuming really takes a good 2 hours to compile)

[javajames]

The cause (i have found now) is killing the last enemy, which crashes the game and puts it into this state. For some reason it can be fixed by remove ~/.local/share/overgrowth

[Faw]

FYI, David escalated this issue to me and I am hot on fixing it.. I actually believe I have it fixed this time.. It's "yet another fluke crash caused by Awesomium". Even though it crashes in angelscript.. it's caused by awesomium.. More specifically it's caused by chromium's use of an "alternate" malloc that seems to randomly be used or not used and causing crashes depending on some random even during linking.

Anyways.. disabling the alternate malloc so far has fixed the issue. I just have to go through the excruciating process of recompiling awesomium on 32bit now. (More time consuming really takes a good 2 hours to compile)

Awesome to hear the cause! Is the 32bit version also going to be running on 64bit then or does that need to be recompiled as well?

[Rhaedas]

I had thought some Linux users with problems were running in 64 bit. At any rate, glad to know the probable cause is located, and maybe we can get the Linux build up and running again. Thanks to those who are working on it!

[Cheeseness]

It's "yet another fluke crash caused by Awesomium".

Glad to hear it's been identified
Keep up the good work <3