[1.16.5 Mixin Crash] Unexpected null pointer exception trying to find ModList.isLoaded when using older generation hardware
Kaleidio opened this issue · 9 comments
The only differences between all other players and this specific person's install is their CPU generation. They are using a 4th gen Core i7-4770, as noted in the log. However, all the other players I had have at least 9th gen or newer and no such crash has ever occurred for them. This seems to be a thread race condition that causes forgery to fail first, then takes down Redstone Control with it in a domino style, something to do with the server config mixin.
tl;dr seems to be an incompatibility with older hardware and Redstone Control, but only if older hardware is present
is there a way I can disable whatever feature would cause your config mixins to fail?
is it a JVM specific bug, should I continue instructing the user to try updating java?
is it a forge bug?
quote of importance:
Caused by: java.lang.NullPointerException
at net.minecraftforge.fml.ModList.isLoaded(ModList.java:181) ~[?:?]
at com.unascribed.fabrication.EarlyAgnos.isModLoaded(EarlyAgnos.java:27) ~[?:3.1.4+1.16]
the patch couldof been released a month or so ago, and probably shouldof.
una does releases and i probably forgot to mention it at the time, and una operates on usa times so probably will be abit before a release.
alrighty, I'll keep tabs on it :) will close the issue once I have a build to test with
this seames like a duplicate of #580 , which was fixed in 3.1.5
which atm is unreleased; if you're in a hurry you can compile the 3.0/1.16 branch (you'll need forgeryTools, which uh you'd also need to compile)
understood. sadly I have to wait for official curseforge release as it's only important for modpack use. Thanks!
as noted by #580 it must be because I have limited_runtime_configs enabled in my ini. I will disable it now to avoid the crash
How soon are we expecting the patch? Attempting to disable the feature supposedly at fault didn't fix the issue for my user. I could try to compile but I'm busy today troubleshooting all the other projects I'm working on too.
is it possible that, unlike the mentioned issue, since I disabled the same config option and it didn't work, that it might be a separate issue?
it's not a conflicting feature.
it's just a part of forge that throws when called too early.
limited_runtime_configs is generally better for compat, since it reduces injections, but it can't do anything in this case.
the issue is the same and was fixed a while ago just never released.
"it's just a part of forge that throws when called too early." - yeah I figured as much I was just making sure. it makes sense as well since forge's modloading order always does mixins before or alongside mod listing
what I mean to ask is there anything left on dev that needs to be done before the build can be uploaded to curse? I won't be able to release the pack update I'm working on until forgery is at least able to load into a world. As I mentioned before it works on my machine, but mine is miles faster than the problem machine