Crash with version 3.4.9.8 for 1.12.2, previous one working fine
Curunir opened this issue ยท 21 comments
Mod Version: DynamicSurroundings-1.12.2-3.4.9.8.jar
Forge Version: 14.23.2.2624
Optifine Version: 1.12.2_HD_U_C9
Link to crash log: https://hastebin.com/ponokiceha.go
Description: This is my personal, private 1.12.2 pack, currently in development. I've been using build 3.4.9.7 until today, then tested with 3.4.9.8. The pack itself loads fine, but the game crashes upon loading the singleplayer test world. I reverted back to 3.4.9.7 and the problem went away. Did not try with freshly generated world yet.
Yeah. Given what I know of these types of errors and how this particular collection is used, it appears there is a thread running "client" side that is not the client thread and it is doing an entity update of some sort. Unfortunately when we get a Minecraft crash report we do not have visibility into the call stacks of the other threads to see what is going on. Upshot is:
- There is a bug somewhere. I don't think it is DS.
- It appears timing related so the chances of it happening may be uncommon.
- We don't know what the other threads are doing. I need to dig into things to see if I can manage to capture that data somehow.
Let me know if I can help. If I find a mod conflicting with yours, that mod will drop out of my pack until it's fixed.
I may make a debug one-off build for you to try. It will have additional diagnostics to ensure the update is occuring on the client thread.
Find a test JAR here. It's the same as 3.4.9.8, but with an additional check to ensure the living update is coming in on the client thread. If for some reason it is not an exception will be generated and your client will crash with the info which hopefully will point a finger at the offender.
For future test builds, I recommend marking it as such both in filename and mod description.
I ran it with the test build, no crashes. Let me know if you have use for any other logs. I'd also be happy to run further tests if you need me to. It would be helpful to identify the bad apple here, since I am running a number of not-so-great mods for testing and could use pointers on which ones to drop out of the pack.
The version thing is on my list of things to do. I may have to make code changes to support parsing out the version string properly.
Assume Bug with Optifine 1.12.2 U C9 I think this is the issue
https://www.youtube.com/watch?v=QUI7iRR9FD8
I haven't looked into the OptiFine JAR that much, so it could be. However, I suspect what is demonstrated in the video has something to do with OptiFine client side chunk caching. If that cache gets screwed up somehow weird rendering things can happen.
The issue has not resurfaced so far. I am keeping an eye on it and will share further crash logs, if any.
Got it to crash the same way again. Happened after I added a few mods and updated others. Again, game started fine, but crashed on loading the worldsave. Unfortunately, I have not been running your modified DS version at the time. Here is the log anyway: https://hastebin.com/uwegulitoy.bash
@Curunir I assume the crash hasn't happened yet? It wouldn't surprise me if it didn't. These types of crashes sometimes require the moons to align.
It hasn't occurred again. I did see your note in the changelog for 3.4.9.13. Good luck getting proper data on it.
@Curunir Have you come across this problem again? I am reading a new release and I am going to activate this check so if anyone else comes across it I can hopefully get a good trace.
@Curunir Have you checked out 3.4.9.14 RE: this issue? In that particular version I changed the method of how DS traverses the map object. I haven't heard anyone encountering the concurrency issue since then.
I will check once I get back to building a pack for 1.12.2. Due to the rapid updates, I took a break and went back to building a stay-behind 1.7.10 Pack.
Closing due to inactivity. If the issue does come back up feel free to re-open.
@Curunir Following up. I was wondering if you had further updates?