[Dev Team] New multiplayer beta + gameplay fixes opt-in patch 1.7.2

[Kaoschan]

I can confirm this
1.7.2 has a high load of pops_api.dll!locale_charset+0x8362 around 12-13% all the time.

 

[Keltosh]

I can confirm this
1.7.2 has a high load of pops_api.dll!locale_charset+0x8362 around 12-13% all the time.

I assume you have an 8-core cpu? It seems to me that pops_api.dll!locale_charset+0x8362 is basically using one core 100%. And if you have an 8-core cpu, your results are consistent with mine, since I have a 4-core cpu

 

[Jervaise]

Hi all,

In order to help us pin down the issue even further there's one thing you can help out with. This is a general tip which is great when trouble shooting these kind of issues and works particularly well when seeing issues with CPU usage. It will require you to be on Windows though.

There's a tool from Microsoft called Process Explorer which can be downloaded here: https://technet.microsoft.com/en-us/sysinternals/processexplorer.aspx

If you unpack and run the executable you'll see a stellaris.exe when you start the game. If you right click that and select Properties you'll get a number of tabs. The one you're interested in is called Threads (it might ask you to download another dll, you can ignore that).

Under the Threads tab you're going to see several parts of the game and the CPU usage of that part. If you see anything that is constantly high, please let me know. For example: if a thread with the start address pdx_mp.dll would constantly max the CPU we would know it's something in the MP layer. If it's Stellaris.exe it could be in the game code etc.

I realize that this is a bit technical, sorry about that

pops_api.dll for me

here are some screenshots:

http://imgur.com/a/EF0Af

 

[Daywalker]

launcher started :
13984 24.93 3.240.742.937 pops_api.dll!locale_charset+0x8362
Main Menu:
13984 24.76 3.223.332.229 pops_api.dll!locale_charset+0x8362
8036 20.39 2.649.832.845 nvd3dum.dll!QueryOglResource+0x1beaef

playing and in Pause:
8036 9.37 1.226.603.792 nvd3dum.dll!QueryOglResource+0x1beaef
13984 24.59 3.205.848.367 pops_api.dll!locale_charset+0x8362
3140 15.35 958.987.197 stellaris.exe!PHYSFS_writeULE64+0x17a12

Stacks for pops_api.dll:
ntoskrnl.exe!memset+0x64a
ntoskrnl.exe!KeWaitForMultipleObjects+0xd52
ntoskrnl.exe!KeWaitForSingleObject+0x19f
ntoskrnl.exe!__misaligned_access+0xbd4
ntoskrnl.exe!__misaligned_access+0x186d
ntoskrnl.exe!IoFreeErrorLogEntry+0x287
pops_api.dll!locale_charset+0xe7aa
pops_api.dll!locale_charset+0xeba7
pops_api.dll!locale_charset+0x1aff
pops_api.dll!locale_charset+0x1c7c
pops_api.dll+0x21dca
pops_api.dll+0x223d0
pops_api.dll!locale_charset+0x82b6
pops_api.dll!locale_charset+0x83de
kernel32.dll!BaseThreadInitThunk+0x12
ntdll.dll!RtlInitializeExceptionChain+0x63
ntdll.dll!RtlInitializeExceptionChain+0x3

https://picload.org/view/ricadoca/pops.png.html

 

[Peterkleber]

Hi all,


I don’t know where to report this, but yesterday together with my friends, we played for about 8 hours. I encountered multiple bugs on that way, but kept on playing.


So, I ran with (1.7.2) [checksum a220] and the mods :

Deutschkorrektur http://steamcommunity.com/sharedfiles/filedetails/?id=681489287

UI Overhaul 1080p http://steamcommunity.com/sharedfiles/filedetails/?id=684509615


After I lost a war, everything went buggy.


To begin with I could select the same 3 scientists repeatedly if one of the research scientists died. Those 3 were selected as unemployed. Unfortunately, i don't have a save directly after that war ended.

Rhizome of Black is currently doing physics, but still I could assign him to other stuff parallel
Rhizome of Cyan was in social and engineering research parallel.
Save 2252


Another thing is that after the war I couldn't build any research or mining stations:

So even though I had the resources it wouldn't let me build a station.
(This got resolved in a later stage of the game, I don't know the cause)
Save 2252

Resolved in Save 2325



Another thing is that I could have more than my cap of 5 planets:

Save 2325


Another bug, which I could not really show in a pic is during a war with the AI " Sidimarus Star Bldc" and others, " Sidimarus Star Bldc" started another war with an AI to the left and won that one. After that " Sidimarus Star Bldc" was no longer in war with us, but had ships in the system "Estreon" and bombed the planet. However, the AI and the federation couldn't either enter the opponents' systems.

Save 2325


A player ("Bösewichte") reported that he couldn't colonize planets even though he had enough minerals and influence. I took control in single player, but then that happened:

Save 2325

I attached the save files

 

[yiome]

Taken while the game is running (not paused)
On a laptop with I7-4700HQ
CPU4 is constantly under full load

 

[Daywalker]

Hi all,


I don’t know where to report this, but yesterday together with my friends, we played for about 8 hours. I encountered multiple bugs on that way, but kept on playing.


So, I ran with (1.7.2) [checksum a220] and the mods :

Deutschkorrektur http://steamcommunity.com/sharedfiles/filedetails/?id=681489287

UI Overhaul 1080p http://steamcommunity.com/sharedfiles/filedetails/?id=684509615


After I lost a war, everything went buggy.


To begin with I could select the same 3 scientists repeatedly if one of the research scientists died. Those 3 were selected as unemployed. Unfortunately, i don't have a save directly after that war ended.
View attachment 272690 View attachment 272691
Rhizome of Black is currently doing physics, but still I could assign him to other stuff parallel
Rhizome of Cyan was in social and engineering research parallel.
Save 2252


Another thing is that after the war I couldn't build any research or mining stations:
View attachment 272692View attachment 272693
So even though I had the resources it wouldn't let me build a station.
(This got resolved in a later stage of the game, I don't know the cause)
Save 2252
View attachment 272696View attachment 272697
Resolved in Save 2325



Another thing is that I could have more than my cap of 5 planets:
View attachment 272694
Save 2325


Another bug, which I could not really show in a pic is during a war with the AI " Sidimarus Star Bldc" and others, " Sidimarus Star Bldc" started another war with an AI to the left and won that one. After that " Sidimarus Star Bldc" was no longer in war with us, but had ships in the system "Estreon" and bombed the planet. However, the AI and the federation couldn't either enter the opponents' systems.
View attachment 272695
Save 2325


A player ("Bösewichte") reported that he couldn't colonize planets even though he had enough minerals and influence. I took control in single player, but then that happened:
View attachment 272698
Save 2325

I attached the save files

the mods are not made for 1.7.2 also the deutschkorrekrtur mod is crap it causes massiv desyncs just remove it and load your save. using a beta with mods not updated for that is kind of useless, of course it will produce bugs

 

[Kliwarrior]

I don’t know where to report this,

In the bug forum

Another thing is that I could have more than my cap of 5 planets:

5 systems , not planets. It seems to me you have 2 planets in the same system (barnard's star).

 

[The Founder]

Yes, it is that pops_api.dll . Its CPU usage is constantly ~24.8%, which I assume means 1 core 100% since my CPU has 4 cores.

Can confirm that file and those percentages (I5, 2 real cores 4 Virtual Cores). The stack looks around like this (I asume from bottom to top):

pops_api.dll!locale_charset+0xeb0d
pops_api.dll!locale_charset+0xebb2
pops_api.dll!locale_charset+0x1aff
pops_api.dll!locale_charset+0x1c7c
pops_api.dll+0x220b8
pops_api.dll+0x223d0
pops_api.dll!locale_charset+0x82b6
pops_api.dll!locale_charset+0x83de
ntdll.dll!RtlInitializeExceptionChain+0x8f
ntdll.dll!RtlInitializeExceptionChain+0x5a

If I read that right, you are somewhere throwing and swallowing exceptions. Wich I consider one of the deadly sins of programming, Chapter Exception handling
Of course deadly Sins of programming like that have to happen with a large Project like Stellaris.

 

[AlphaAsh]

Yes, it is that pops_api.dll . Its CPU usage is constantly ~24.8%, which I assume means 1 core 100% since my CPU has 4 cores.

Confirmed here too. Something to do with pops recently changed is running away every tick. Suspects off the top of my head - checking whether to auto-migrate, recent change to faction attractors, some interconnectivity with the new hard-code to prevent rampant terraforming. I'm going to comment out a few of the recent soft-code changes related to those (if any) and see if that helps.

 

[GAGA Extrem]

So pops are exploiting the game at the expense of the player?
...sounds like Vicky 2 to me.

Anyway, I'd guess that the Stellaris team will be able to get to wörk on that once the long weekend is over.

 

[Dodge]

How dare the pops get brains of their own, the humanity

 

[Labatyd]

I have the exact same problem.

Same problem for me. 1.7.2. does start the launcher but not the game.

 

[Risa]

I'm going to comment out a few of the recent soft-code changes related to those (if any) and see if that helps.

Can't be soft-code related. That thread uses a full core in launcher, at which point (almost) none soft code is loaded yet.

 

[perost]

I tried 1.7.2 and saw the same issue on Linux, the launcher is using 100% of one core while previously it only used a few percent. I ran it through callgrind (part of valgrind) for a quick profiling, and this cycle used ~90% of the CPU time:

This is just the launcher mind you, but trying to run the game through valgrind is beyond my patience (it took several minutes for the launcher to even appear). And from other reports it seems like this issue affects both the game and the launcher.

 

[Ex Mudder]

Memory leak, runaway process, infinite loop?

INFINITY MACHINE?

Or STRANGE LOOP?

I'm putting my money on the Contingency.

 

[grandad1982]

Like others I have stuttering where I didn't before. I have a nearly new skylake i5, 16gb of ddr4, gtx 970 and its on a Samsung evo ssd. Drivers are current.

 

[AlphaAsh]

Can't be soft-code related. That thread uses a full core in launcher, at which point (almost) none soft code is loaded yet.

True. Also weird that that dll has anything to do with the launcher anyway. So new suspect, save-game verifier/updater.

edit: Possible test-case: Backup saves, clear out save folder. Run game, start new game. Potentially if there's no save-games to verify or update, runaway function has no data to work on.

edit 2: Which leads to suspect number 2, and someone else in this thread hinted at it as well. Steam cloud synchronization.

edit 3: Be interesting to get some stats on who is seeing the problem and has clound synching on, and who isn't and has cloud synching off.

 

[macaroni_salad]

ETHAN BRADBURY

 

[AlphaAsh]

Initial testing is good here, although with ambient temperatures being a lot lower today, that might be throwing the test results.

Without cloud synching, the launcher went mental on core 6 but that settled after starting the game. I've now been continuing a busy mid-game save, and the CPU fan has not ramped up at all. I'm even alt-tabbed out of fullscreen atm, and still no significant CPU load.

So try testing without cloud synching saves all, get some results for comparison.