View Issue Details
ID | Project | Category | Date Submitted | Last Update | |
---|---|---|---|---|---|
0003872 | AI War 1 / Classic | Crash/Exception | Sep 20, 2011 6:07 pm | Jan 30, 2012 8:10 pm | |
Reporter | TechSY730 | Assigned To | Chris_McElligottPark | ||
Status | closed | Resolution | unable to reproduce | ||
Product Version | 5.017 | ||||
Summary | 0003872: Fatal Error in GC | ||||
Description | I am sometimes getting a crash accompanied with a message box with the title "Fatal Error in GC", and message "GetThreadContext failed" Nothing from that crash is in the unhanded errors file. OS: Ubuntu Linux 11.04 (kernel version 2.6.38-11-generic) Wine 1.2.3 using Windows XP mode OK, I realize this is not a supported configuration, so I may be on my own, but I am wondering if others are getting this message. | ||||
Tags | No tags attached. | ||||
Internal Weight | New | ||||
related to | 0005059 | resolved | Chris_McElligottPark | Valley 1 | Fatal Error in GC |
|
I remember that GetThreadContext funnybusiness from back when we were doing the Unity port. I just forget what it was about. Possibly about how on the Mac version the Mac OSX wouldn't let us check to see if another copy of the process was running to see if we needed to not do certain networking things. |
|
Update on the issue. It seems to happen most often when switching views from one planet to another. |
|
Due to its similarity to the AVWW wind crash, I'll answer the same question here. In this game, it seems much more varied about when the crash happens. Typically its after like 2-3 hours before I get this crash, but one time it was a mere 30 minutes before it happened. |
|
Thanks! * Previously it was still possible for the mono garbage collector to suicide itself during save operations if it had to resize a large buffer and the transient memory happened to be too low at the time. ** At core, this is a bug in the way that unity 3d handles lack of heap space, and it's something that we've had to work around with all our games during memory-intensive operations (rather than just garbage collecting automatically any choosing to get on its day -- which it could easily do -- the engine instead throws a fatal error and dies if you use more than 800MB of heap space). ** Preventative code has now been put in place to prevent these sorts of errors during buffer resizings, basically forcing a GC collection if there is more than 500MB of heap currently in use just to be on the safe side. |
|
Just a heads up. This change has not been added to AI War 5.021 release notes yet. |
|
Wasn't planning on putting this in AI War. AI War has its own ways of serializing data which are completely unlike those of AVWW. It also already has its own GC handling to prevent this sort of issue whenever possible. To my knowledge it's working very well except when players have really super extreme scenarios. |
|
Then you probably shouldn't mark this issue as resolved, as this is the copy of the issue for AI War. The related issue is the one for AVWW which you have fixed. |
|
Oh, whoops, I see. Man I keep missing that. Well, I can mark it as "unable to reproduce" if you like. I think your OS really is a factor. |
|
Yea, status of "feedback" or "closed" with "unable to reproduce" as the resolution seems like the way to go. I guess this will have to wait around until someone with a supported configuration gets GC problems as well. Or worse, this may be a Wine bug. I can't use the latest beta's of Wine as even though they fix so many things, they have introduced a bug with mouse input. Though it has been fixed for normal mice, it still is affecting touchpads, which I use. (Yes, I play all my games on touchpad. It's not as bad as you might think.) If they ever fix it, I will update Wine and try to see if this error comes up again. |
|
Given that we have a LOT of people playing both AVWW and AI War these days, and you're playing on an unsupported platform and running into these issues with greater frequency compared to anyone else (on AVWW) and in general when others are not (AI War), I'd have to put my money on it being some form of WINE bug. Unfortunately. :( |
|
Good news, the latest versions of Wine fix both the touchpad input things and GREATLY improves the audio subsystem's implementation. So far, under this new version, no GC problems. |
|
Woohoo! That is indeed really excellent news. :) |
Date Modified | Username | Field | Change |
---|---|---|---|
Sep 20, 2011 6:07 pm | TechSY730 | New Issue | |
Sep 20, 2011 6:51 pm | keith.lamothe | Note Added: 0013260 | |
Oct 17, 2011 12:05 am | TechSY730 | Note Added: 0016629 | |
Oct 27, 2011 11:25 pm | TechSY730 | Relationship added | related to 0005059 |
Oct 28, 2011 9:43 am | TechSY730 | Note Added: 0017278 | |
Nov 6, 2011 12:28 pm | Chris_McElligottPark | Internal Weight | => New |
Nov 6, 2011 12:28 pm | Chris_McElligottPark | Note Added: 0017449 | |
Nov 6, 2011 12:28 pm | Chris_McElligottPark | Status | new => resolved |
Nov 6, 2011 12:28 pm | Chris_McElligottPark | Fixed in Version | => 5.021 |
Nov 6, 2011 12:28 pm | Chris_McElligottPark | Resolution | open => fixed |
Nov 6, 2011 12:28 pm | Chris_McElligottPark | Assigned To | => Chris_McElligottPark |
Nov 8, 2011 2:03 pm | TechSY730 | Note Added: 0017460 | |
Nov 8, 2011 2:14 pm | Chris_McElligottPark | Note Added: 0017461 | |
Nov 8, 2011 2:25 pm | TechSY730 | Note Added: 0017463 | |
Nov 8, 2011 2:28 pm | Chris_McElligottPark | Note Added: 0017464 | |
Nov 8, 2011 2:37 pm | TechSY730 | Note Added: 0017466 | |
Nov 8, 2011 2:44 pm | Chris_McElligottPark | Note Added: 0017467 | |
Nov 8, 2011 2:44 pm | Chris_McElligottPark | Status | resolved => closed |
Nov 8, 2011 2:44 pm | Chris_McElligottPark | Resolution | fixed => unable to reproduce |
Nov 8, 2011 2:44 pm | Chris_McElligottPark | Fixed in Version | 5.021 => |
Jan 30, 2012 8:09 pm | TechSY730 | Note Added: 0018431 | |
Jan 30, 2012 8:10 pm | Chris_McElligottPark | Note Added: 0018432 | |
Apr 14, 2014 9:27 am | Chris_McElligottPark | Category | Bug - Crash or Exception => Crash/Exception |