Weapon.dll crash on dedicated server.

[Shadow KT]

I believe this is the same old culprit, from a thread I had made couple of months ago.I have narrowed down the issue to the BMP-2 vehicle and created a massive battle with over 100 vs 100 BMP-2s fighting, with me jumping in and out of random vehicles, shooting, giving them command to move, etc.

 

I played the mission in SP earlier and it did not crash, but on the dedicated server it does. It took around 5-6 minutes.

 

Log files from the server uploaded. By reading them, can anyone tell me what exactly is the problem ? Is it the ATGM on the BMP-2 ? The damage model of the vehicle ?

dcs.log-20200930-195505.zip

Edited September 30, 2020 by Shadow KT

[Flappie]

Hi Shadow.

 

Oddly enough, your crash doesn't seem to be linked to BMPs or ATGMs but to Vikhrs.

I don't have a clue what "wArmour_scheme::set_ws" is, but the function that called it is named "wAmmunitionVikhr::Ny".

 

Now, I don't see Ka-50 nor Su25T in your mission. Is one of these modules in? Can you please attach the track or at least the mission?

 

2020-09-30 19:55:02.858 INFO Scripting: event:weapon=AT-5 Spandrel,t=29212.26,type=shot,initiatorMissionID=303,

2020-09-30 19:55:03.075 INFO Scripting: event:weapon=AT-5 Spandrel,t=29212.473,type=shot,initiatorMissionID=345,

2020-09-30 19:55:03.270 INFO Scripting: event:weapon=AT-5 Spandrel,t=29212.674,type=shot,initiatorMissionID=343,

2020-09-30 19:55:03.732 INFO Scripting: event:weapon=AT-5 Spandrel,t=29213.118,type=shot,initiatorMissionID=302,

2020-09-30 19:55:04.277 INFO EDCORE: try to write dump information

2020-09-30 19:55:04.295 INFO EDCORE: # -------------- 20200930-195505 --------------

2020-09-30 19:55:04.302 INFO EDCORE: DCS/2.5.6.55743 (x86_64; Windows NT 10.0.17763)

2020-09-30 19:55:04.309 INFO EDCORE: C:\Program Files\Eagle Dynamics\DCS World OpenBeta Server\bin\Weapons.dll

2020-09-30 19:55:04.315 INFO EDCORE: # C0000005 ACCESS_VIOLATION at 9E57F4BD 00:00000000

2020-09-30 19:55:04.363 INFO EDCORE: SymInit: Symbol-SearchPath: '.;C:\Program Files\Eagle Dynamics\DCS World OpenBeta Server;C:\Program Files\Eagle Dynamics\DCS World OpenBeta Server\bin;C:\windows;C:\windows\system32;SRV*C:\websymbols*http://msdl.microsoft.com/download/symbols;', symOptions: 530, UserName: 'DCS.Dedicated.Admin'

2020-09-30 19:55:04.370 INFO EDCORE: OS-Version: 10.0.17763 () 0x100-0x1

2020-09-30 19:55:05.890 INFO EDCORE: 0x000000000009F4BD (Weapons): wArmour_scheme::set_ws + 0xB2D

2020-09-30 19:55:05.897 INFO EDCORE: 0x00000000000950DA (Weapons): wAmmunitionVikhr::Ny + 0x514A

2020-09-30 19:55:05.909 INFO EDCORE: 0x0000000000092BC5 (Weapons): wAmmunitionVikhr::Ny + 0x2C35

2020-09-30 19:55:05.914 INFO EDCORE: 0x00000000000C2873 (Weapons): wWeaponSystem::getDiffInitStatus + 0x513

2020-09-30 19:55:05.920 INFO EDCORE: 0x00000000000D454D (Weapons): wWeaponSystemManager::getDiffInitStatus + 0x5D

2020-09-30 19:55:05.926 INFO EDCORE: 0x000000000008467D (Transport): woCar::useMarkedSmoke + 0x62CD

2020-09-30 19:55:05.932 INFO EDCORE: 0x0000000000078C43 (Transport): woCar::onNetHumanTakeControl + 0xF3

2020-09-30 19:55:05.938 INFO EDCORE: 0x00000000000826BF (Transport): woCar::useMarkedSmoke + 0x430F

2020-09-30 19:55:05.944 INFO EDCORE: 0x0000000000103F83 (edCore): Mail::Reader::stream + 0x1473

2020-09-30 19:55:05.951 INFO EDCORE: 0x000000000089851F (DCS): CoreUtils::TempFilesManager::operator= + 0x52A8EF

2020-09-30 19:55:05.958 INFO EDCORE: 0x0000000000898AA1 (DCS): CoreUtils::TempFilesManager::operator= + 0x52AE71

2020-09-30 19:55:05.964 INFO EDCORE: 0x00000000008AC43B (DCS): CoreUtils::TempFilesManager::operator= + 0x53E80B

2020-09-30 19:55:05.971 INFO EDCORE: 0x0000000000888A0E (DCS): CoreUtils::TempFilesManager::operator= + 0x51ADDE

2020-09-30 19:55:05.977 INFO EDCORE: 0x0000000000878212 (DCS): CoreUtils::TempFilesManager::operator= + 0x50A5E2

2020-09-30 19:55:05.983 INFO EDCORE: 0x0000000000724354 (DCS): CoreUtils::TempFilesManager::operator= + 0x3B6724

2020-09-30 19:55:05.989 INFO EDCORE: 0x00000000006F6604 (DCS): CoreUtils::TempFilesManager::operator= + 0x3889D4

2020-09-30 19:55:05.995 INFO EDCORE: 0x00000000006F69C9 (DCS): CoreUtils::TempFilesManager::operator= + 0x388D99

2020-09-30 19:55:06.001 INFO EDCORE: 0x00000000016E2CC4 (DCS): AmdPowerXpressRequestHighPerformance + 0xB14CC0

2020-09-30 19:55:06.007 INFO EDCORE: 0x00000000008E261E (DCS): CoreUtils::TempFilesManager::operator= + 0x5749EE

2020-09-30 19:55:06.013 INFO EDCORE: 0x0000000000017974 (KERNEL32): BaseThreadInitThunk + 0x14

2020-09-30 19:55:06.019 INFO EDCORE: 0x000000000006A271 (ntdll): RtlUserThreadStart + 0x21

2020-09-30 19:55:08.241 INFO EDCORE: Minidump created.

There are some precedents with this faulty wArmour_scheme::set_ws function:

 

 

https://forums.eagle.ru/showthread.php?t=284754

https://forums.eagle.ru/showthread.php?t=273592 (your previous post about this error)

https://forums.eagle.ru/showthread.php?t=271627

Edited September 30, 2020 by Flappie

[Shadow KT]

I repeated the tests with BMP-3 and M2A2s and did not have a crash. I added even more vehicles fighting.... fps got extremely low, especially in the BMP-3 battle, but no crash.

 

No aircraft at all on the mission. Only other units on the mission are a column of different vehicles , which are far away and just idling.

 

You can find the server track in the zip, but I am also uploading the client one.

 

If I/we can gather a bigger group to interact with the vehicles, we may be able to crash the server even faster. I don't know why it calls out a Vikhr, maybe it is some old code, which has never been edited. I guess it checking the damage model/armor scheme on interaction with the called waepon and it crashes for some reason... just guessing here, ED coders will have to answer that mystery.

 

Sometimes it just feel like the first time you launch DCS after starting your machine, it just needs to get its drama out and crash, before it can run normally.... Repeated the BMP-2 test twice more and wasn't able to create a new crash.

BMP-2 Test-20200930-224935.trk

Edited September 30, 2020 by Shadow KT

[Shadow KT]

I played both server and client track over and I did not get a crash. As I noted in my original post, can't seem to get a crash in SP.

 

Don't know what magic developers gotta do, but I think tracks and logs can be correlated so the issue can be found.

 

This really sucks, especially when you are creating and playing in-depth missions, which require people to carefully plan and move units, just to have the game crash minutes.

 

It has to be unit related, as I've seen many other servers, with hundreds of units and multiple scripts running (I don't run scripts) have vehicles fight, but not crash

Edited September 30, 2020 by Shadow KT

[Flappie]

Even in SP, your mission was barely playable on my client (2600X, 32 GB RAM, 1080 Ti). I'm not sure DCS is quite ready for huge tank battles.

[Shadow KT]

You are missing the point.... this is an exaggerated setup, so an issue can be shown.

 

If it was in normal circumstances, then we will get no where as I've already done such reports and the answer is.... your track is too long, so meh...

 

This does happen in normal conditions too, it just takes time. It might happen an hour in or two hours into a mission. Easiest way to record the issue is to force it. No it is not related to numbers either, this can happen with 50 units or 500 units... all the same, the chances of it happening are just less.

 

All my tests ran fine and were playable. They ran even better on my own PC in SP. Don't know what issues you had, but it was running fine on a 3570K with a 1070 non Ti

Edited October 1, 2020 by Shadow KT

[BIGNEWY]

Im not able to reproduce this yet sorry, I have tried

 

thanks

[Shadow KT]

Okey.... can we organize a meeting or something so we can all get on a server and crash it ?

[Flappie]

You are missing the point.... this is an exaggerated setup, so an issue can be shown.

Got you. :thumbup:

 

 

Okey.... can we organize a meeting or something so we can all get on a server and crash it ?

I'm OK. I've sent you an MP.