GridArea<std::shared_ptr<LevelChunk>>::move related crash

[wcdfilll]

i only have an image about bug, it isn't feedback error anything. What I want to ask is, what could it be due to? Port, Plugin or something ? I need to narrow down the error so that I can judge and fix it completely. If it is difficult to judge the error, I will provide more pictures

[KawaiiNahida]

Every time you execute the command "/home go" BDS will crash immediately?

[wcdfilll]

that's not it, even if it's not home add or del, it's still interrupted unexpectedly. I can send the file server over to have a better view

[KawaiiNahida]

Can you use a debugger(x64dbg, windbg, VS) or enable the builtin CrashLogger?

[KawaiiNahida]

We have no idea What Happened, and we have tested that "/home go" command won't make the server crash

[wcdfilll]

no, it's not due to /home, I'm sure it is. I keep it open for a while, and let it hang and then it shuts off on its own. The strange thing is, the player's data (each player's item inventory) does not lose any items. Sometimes they even get duplicate items.

[wcdfilll]

Can you use a debugger(x64dbg, windbg, VS) or enable the builtin CrashLogger?

if this can respond to an error I will send it to you as soon as possible.

[dreamguxiang]

LL2.0.4?

[wcdfilll]

yes

[wcdfilll]

Can you use a debugger(x64dbg, windbg, VS) or enable the builtin CrashLogger?

this is the error after i enable crash logger

[wcdfilll]

I have modified LLhelper's config again.

[dreamguxiang]

Determined the problem

[dreamguxiang]

There is a problem with lltpa, we will update the new version as soon as possible

[wcdfilll]

Determined the problem

but it also has LxL 's fault

[dreamguxiang]

The problem has been fixed, please update lless 2.0.4,
https://github.com/LiteLDev/LiteLoaderPlugins/releases/tag/2.0.4

[wcdfilll]

2.0.4.1:

[wcdfilll]

still same error, I will update the error checking information as soon as possible to fix it

[dreamguxiang]

What did the player do before the server crash

[dreamguxiang]

What did the player do before the server crash

Do you know, or what instructions were executed

[wcdfilll]

What did the player do before the server crash

Do you know, or what instructions were executed

This is the last log file I have. Since there were 15 or more people, I couldn't check them al
CrashDump_20211225_23-20-11.zip
l

[dreamguxiang]

Is your lltpa version is 211229?There is a person who has the same crash log as you, and it has been resolved

[wcdfilll]

CrashDump_20211225_23-20-11.zip
tpa has a version, where can I see it? 211229 ?

[dreamguxiang]

/plugins

[wcdfilll]

I have updated to the latest

[wcdfilll]

I can't try again right now, I'll update again tomorrow

[dreamguxiang]

211229 mainly fixes the crash of the player's no death to use /back.Its crash log is the same as yours.If it can't be solved, it may be due to internal reasons in bds, or there is an error in the data of a player.

[wcdfilll]

So how do you fix it? I add LLtpa but not /back ever, but /tpa often. I don't expect a single player on my map with the wrong data to cause the server to crash

[dreamguxiang]

My friend and I tried all the features of lltpa, but found no bugs, including offline detection of tpa player.If you have the conditions, you can try to run a test server to find a better way to reproduce.The generated crash log, it is completely impossible to see why the place crashed.Except for Level::tick of ll and lxl HOOK, there is no plug-in related information.The method of reproduction is very important, otherwise no one can confirm the location of the crash.

[dreamguxiang]

I can tell you a few characteristics,

1. After the player join the server, the server crashes
2. The player client crashes,
3. The server feigned death and then crashed

We have asked other people, if they have the same problem, allows us to solve it faster

If it's just a case, we may not be able to solve it, because my BDS without plugins loader has also crashed and suspended animation.

[dreamguxiang]

I tried the old version of LLtpa again,/back problem did not cause a crash, it would only cause the executing player to get stuck, and rejoin the server to recover.

I also saw a similar crash on Mojira, which is likely to be the cause of BDS itself

[KawaiiNahida]

I suggect that you can try remove LLHelper and LXL temporary and check if it's Okay, I think that vector<LevelChunk>::resize has a stack overflow issue, which is cause by BDS it's own

[wcdfilll]

I'm thinking of removing lxl first, but if it's ll helper I'll check them one by one if this still happens

[wcdfilll]

211229 mainly fixes the crash of the player's no death to use /back.Its crash log is the same as yours.If it can't be solved, it may be due to internal reasons in bds, or there is an error in the data of a player.

[wcdfilll]

It's been 3 hours since I quit LLhelper. No more crash messages

[dreamguxiang]

It's been 3 hours since I quit LLhelper. No more crash messages

I will check llhelper, wait a minute

[dreamguxiang]

Did you use this feature? try remove it and see if it still crashed randomly

[wcdfilll]

After 9h server open, server crashed once more. This time, i forget enable builtin CrashLogger so i haven't picture about bug. LLhelper just a part. I plan remove LxL

[wcdfilll]

bug still take place. haizzz

[dreamguxiang]

The only case the same as yours was solved by deleting the data of a certain player， Judging from the crash log, it may be that there is a problem with the pos data of a certain player

[wcdfilll]

if so this is very difficult, when deleting a player data with 7000 different account id in one server. Is it possible ?

[KawaiiNahida]

We also recived a lot of report from Chinese Community too, but we never succees in reproducing it. This problem looks wired, most of the dump file and TrackBack show us that this is caused by bds realloc or resize std::vector<std::sharedptr>.
so this might related to bad mcworld or player data which may cause it crashed.
It might also because some plugin hook the level::tick and cause this bug and it seems that this problem doesn't exitst in original bds

[LazuliKao]

I had tested HorionHack half a year ago and tested the CrashServer function
This stack information looks the same.
especially these:GridArea<std: : shared ptr<LevelChunk> >::,ChunkViewSource: : move

it looks like there are hackers on your server.

[dreamguxiang]

You can turn on this option to see if the problem still exist

[dreamguxiang]

We still think this is a crash caused by someone on your server using Horion's crashserver, and turning on BDS's correct-player-movement can effectively prevent it.

[RimuruChan]

Is it still happening to your server？

[dreamguxiang]

If this phenomenon persists, please reopen