Fog Ridl

Hey there, stumbled upon a stack overflow error using the spine-corona runtime.

Basically, leaving the game open for several hours will eventually lead to a crash (the number of hours varies for reasons unknown to me). The log references spine.lua functions making me suspicious of the Spine runtime immediately, but to make sure I tested my current project running Spine vs. an older one that didn't, and only the one using Spine crashed.

Out of curiosity, I downloaded the latest example project from the Spine runtimes github (https://github.com/EsotericSoftware/spine-runtimes/tree/3.8/spine-corona) and tested again without making any changes to the code — the same crash happens in this official default configuration as well.

Here's the log:

Test start (for timeframe reference):
03:02:15.148
03:02:15.148 Corona Simulator 2019.3548 (Nov 7 2019 04:29:56)
03:02:15.148
03:02:22.464
03:02:22.464 Copyright (C) 2009-2019 C o r o n a L a b s I n c .
03:02:22.464 Version: 3.0.0
03:02:22.464 Build: 2019.3548
03:02:22.464 Platform: GenericAndroidDevice / x64 / 10.0 / Intel(R) UHD Graphics 630 / 4.5.0 - Build 25.20.100.6793 / 2019.3548 / en_US | US | en_US | en
03:02:22.464 Loading project from: C:\Users\fogridl\corona\spine-corona
Stack overflow:
05:19:20.503 ERROR: Runtime error
05:19:20.503 ?:0: C stack overflow
05:19:20.503 stack traceback:
05:19:20.503 ?: in function 'cloneArray'
05:19:20.503 ?: in function <?:205>
05:19:20.503 [C]: ?
05:19:20.503 C:\Users\fogridl\corona\spine-corona\spine-corona\spine.lua:248: in function 'flush'
05:19:20.503 C:\Users\fogridl\corona\spine-corona\spine-corona\spine.lua:208: in function 'updateWorldTransform'
05:19:20.503 C:\Users\fogridl\corona\spine-corona\main.lua:177: in function '?'
05:19:20.503 ?: in function <?:205>
05:19:20.503 [C]: ?
05:19:20.503 C:\Users\fogridl\corona\spine-corona\spine-corona\spine.lua:248: in function 'flush'
05:19:20.503 C:\Users\fogridl\corona\spine-corona\spine-corona\spine.lua:208: in function 'updateWorldTransform'
05:19:20.503 ...
05:19:20.503 [C]: ?
05:19:20.503 C:\Users\fogridl\corona\spine-corona\spine-corona\spine.lua:248: in function 'flush'
05:19:20.503 C:\Users\fogridl\corona\spine-corona\spine-corona\spine.lua:208: in function 'updateWorldTransform'
05:19:20.503 C:\Users\fogridl\corona\spine-corona\main.lua:177: in function '?'
05:19:20.503 ?: in function <?:205>
05:19:20.503 [C]: ?
05:19:20.503 C:\Users\fogridl\corona\spine-corona\spine-corona\spine.lua:248: in function 'flush'
05:19:20.503 C:\Users\fogridl\corona\spine-corona\spine-corona\spine.lua:208: in function 'updateWorldTransform'
05:19:20.503 C:\Users\fogridl\corona\spine-corona\main.lua:177: in function '?'
05:19:20.503 ?: in function <?:205>
Would really love a fix for this, ~2 hours is sadly too fast to say it's tolerable and not too important. Any temporary workarounds are also welcome if you can suggest any.

Thanks in advance!
Fog Ridl
  • 記事: 2

Mario

Huh, that is amazingly weird. The spine-corona and spine-lua runtime don't really do anything that could cause a stack overflow. Based on the logs, it seems this happens on the desktop. I'll try to repro!
アバター
Mario

Mario
  • 記事: 2236

Fog Ridl

badlogic さんが書きました:Huh, that is amazingly weird. The spine-corona and spine-lua runtime don't really do anything that could cause a stack overflow. Based on the logs, it seems this happens on the desktop. I'll try to repro!
Have you been able to reproduce the issue? Sometimes it happens sooner, sometimes later (about 10 hours was my max if I'm not mistaken), but it happens pretty much always in my case. Maybe there's some additional info I could provide. You got that right though, it happens on the desktop (Win 10, x64).
Fog Ridl
  • 記事: 2

Mario

I have not been able to reproduce it, I've only tested on macOS so far. I'll give it a whirl on Windows, maybe there's a difference in the Lua interpreter there. I just don't understand how this specific call trace could ever produce a stack overflow.
アバター
Mario

Mario
  • 記事: 2236


Return to Runtimes