Prepar3D crashing with 4.949f

[Oliver Grützmann]

Hi,

 

since installing FSUIPC 4.949f I get Prepar3D crashes in every flight referencing ntdll.dll.

 

It might be related to FSUIPC, in fact it started just after installing 4.949f (I was doing a 2.5 hrs flight without problems, shutting down my PC, out for a walk, starting my PC, installing .949f and doing a flight).

It's a hard CTD, so there's no hint in the log file. The crash happens randomly (first flight it was after 30 minutes, second after 1.5 hrs).

Really no changes but FSUIPC, same programs running (FS Global Real Weather and FS RealTime via widefs).

 

Could this be caused by the changes in the lua code? I'm using DynamicFriction.lua

Any ideas?

 

Oliver.

 

Here's the log of FSUIPC and the appcrash itself:

********* FSUIPC4, Version 4.949f by Pete Dowson *********
Prepar3D.exe version = 3.1.2.15831
Reading options from "D:\Prepar3D v3\Modules\FSUIPC4.ini"
Running inside Prepar3D v3 on Windows 8.0
Module base=57CE0000
FSUIPC4 Key is provided
WideFS7 Key is provided
      312 System time = 14/02/2016 22:44:59
      328 FLT UNC path = "\\VIRUS\Users\Olli\Documents\Prepar3D v3 Files\"
      328 ------ Module Version Check ------
      328        acontain.dll: 3.1.2.15831
      328             api.dll: 3.1.2.15831
      328        controls.dll: 3.1.2.15831
      328      fs-traffic.dll: 3.1.2.15831
      328             G3D.dll: 3.1.2.15831
      328        language.dll: 3.1.2.15831
      328            sim1.dll: 3.1.2.15831
      328        visualfx.dll: 3.1.2.15831
      328         weather.dll: 3.1.2.15831
      328          window.dll: 3.1.2.15831
      328 ----------------------------------
      328 Trying D:\Prepar3D v3\Modules\SimConnectP3D2.dll
      328 Found it: trying to connect
      344 FS UNC path = "\\VIRUS\Prepar3D v3\"
      500 ---------------------- Joystick Device Scan -----------------------
      500 Product= Saitek X52 Pro Flight Control System
      500    Manufacturer= Saitek
      500    Vendor=06A3, Product=0762 (Version 0.48)
      500    Serial Number= Saitek
      500 -------------------------------------------------------------------
      515 LogOptions=00000000 00000001
      515 -------------------------------------------------------------------
      515 ------ Setting the hooks and direct calls into the simulator ------
      515 --- CONTROLS timer memory location obtained ok
      515 --- SIM1 Frictions access gained
      515 --- FS Controls Table located ok
      515 --- Installed Mouse Macro hooks ok.
      515 --- Wind smoothing fix is installed
      515 --- SimConnect intercept for texts and menus installed ok
      515 --- All links okay (except older global weather setting method)
      515 -------------------------------------------------------------------
      515 SimConnect_Open succeeded: waiting to check version okay
      515 Trying to use SimConnect Prepar3D
    39687 Running in "Lockheed Martin® Prepar3D® v3", Version: 3.1.3.1 (SimConnect: 3.1.0.0)
    39687 Initialising SimConnect data requests now
    39687 FSUIPC Menu entry added
    39703 \\VIRUS\Users\Olli\AppData\Local\Lockheed Martin\Prepar3D v3\Prepar3D_Default.fxml
    39703 \\VIRUS\Prepar3D v3\SimObjects\Airplanes\IRIS Raptor Driver\Raptor.air
    59265 Weather Mode now = Theme
    72781 \\VIRUS\Prepar3D v3\SimObjects\Airplanes\PMDG 737-800NGX WL\B737-800WL.air
    86797 System time = 14/02/2016 22:46:25, Simulator time = 08:00:00 (13:00Z)
    86812 Aircraft="Boeing 737-8K5NGX airberlin Winglets"
    87797 Starting everything now ...
    91297 Advanced Weather Interface Enabled
   299297 Weather Mode now = Global
   299297 Weather Mode now = Custom
   906500 Weather Mode now = Global
   906500 Weather Mode now = Custom
  3187390 Sim stopped: average frame rate for last 3012 secs = 47.9 fps
  3407265 Sim stopped: average frame rate for last 209 secs = 59.7 fps
  3650390 Sim stopped: average frame rate for last 240 secs = 59.6 fps

Name der fehlerhaften Anwendung: Prepar3D.exe, Version: 3.1.2.15831, Zeitstempel: 0x5672b60a
Name des fehlerhaften Moduls: ntdll.dll, Version: 10.0.10586.103, Zeitstempel: 0x56a853dc
Ausnahmecode: 0xc0000374
Fehleroffset: 0x000dc759
ID des fehlerhaften Prozesses: 0x1ebc
Startzeit der fehlerhaften Anwendung: 0x01d16770ee6604ac
Pfad der fehlerhaften Anwendung: D:\Prepar3D v3\Prepar3D.exe
Pfad des fehlerhaften Moduls: C:\Windows\SYSTEM32\ntdll.dll

[Pete Dowson]

since installing FSUIPC 4.949f I get Prepar3D crashes in every flight referencing ntdll.dll.

 

It might be related to FSUIPC, in fact it started just after installing 4.949f (I was doing a 2.5 hrs flight without problems, shutting down my PC, out for a walk, starting my PC, installing .949f and doing a flight).

 

And what was the previous version of FSUIPC you were using. And what other things have you changed?

 

It's a hard CTD, so there's no hint in the log file. The crash happens randomly (first flight it was after 30 minutes, second after 1.5 hrs).

Really no changes but FSUIPC, same programs running (FS Global Real Weather and FS RealTime via widefs).

 

Could this be caused by the changes in the lua code? I'm using DynamicFriction.lua

 

FSUIPC is very unlikely to be causing any sort of crash in NTDLL itself. There most certainly haven't been any changes in Lua which would affect the use of DynamicFriction.lua.

 

I do think there is a particular problem in the SimConnect part of P3D3.1 which does cause NTDLL crashes, and it may simply be that the slightly different memory arrangements you get with different versions of any add-in create a crash in situations where otherwise it would simply be some sort of data glitch.

 

There have been quite a few reports of crashes in NTDLL on the L-M forum. Mostly they tend to be access violations (error 5). Yours is error 0x374, which doesn't even appear to be listed in the usual MSDN list of codes.  The only possibly lead I found using google is https://technet.microsoft.com/en-us/library/hh859221.aspx which appears to be talking about the file system and an error trying to delete something. Could you be using FSUIPC's autosave at all? Else I don't think FSUIPC specifically deletes files in normal operation.

 

I suggest that you add your report to those already on the L-M forum.

 

Regards

Pete

[Oliver Grützmann]

He Pete,

 

thanks for your  reply. FSUIPC was my first guess, it started just after the upgrade.

 

 

And what was the previous version of FSUIPC you were using. And what other things have you changed?

 

Previous was 4.949c or later, not sure if there was a version in between. I checked last monday and there was no update available. As mentioned, nothing else changed but FSUIPC. No new scenery, no other modules, no settings, no cfg tweaks, no Windows updates, nothing. I was shutting down my computer after one flight, booted later on, installed FSUIPC .949f and started a flight with the mentioned result.

I had a rock solid P3D before, no crashes at all.

 

 

The only possibly lead I found using google is https://technet.micr...y/hh859221.aspx which appears to be talking about the file system and an error trying to delete something. Could you be using FSUIPC's autosave at all? Else I don't think FSUIPC specifically deletes files in normal operation.

No, I'm not using autosave. Had no use for it, since I never experienced any crashes or OOMs.

 

Is there any chance to get back to 4.949c to start sorting out the error? At least I could rule out or confirm that it's in some way related to it (and, if so, at least get back to a stable sim).

 

I'm at work currently, but I'll do some testing back at home. What I found in LM forums was someone mentioning saitek drivers. Maybe it is somehow related to this and the upgrade. Have to admit, I'm completely clueless - from absolutely stable to constantly crashing within a few hours.

 

Oliver.

[Pete Dowson]

Previous was 4.949c or later

 

Well there are only minor changes between c and f, namely:

 

1. Correction to erroneous error message when assigning buttons

2. Extra logging for ASN WXreadar file problems

3. Fixes for errors in Lua functios to do with setting and clearing bits in values

4. A new Lua ext library function, "hasfocus"

 

and that's it. All those changes only do anything if you are using the mentioned items, no changes whatsoever have been made to anything you would be using in normal running.

 

As I said, sometimes when there is actually a problem occurring, you only actually get a resulting crash when the memory layout changes a little. If the update to FSUIPC really is the change which precipitated the result you are getting then I do think it's a pre-existing problem in any case which simply hasn't before resulted in a crash.

 

Is there any chance to get back to 4.949c to start sorting out the error?

 

You'll need to give me your email address. I'm not uploading one. But please also read what I'm saying above.

There was an "e" release too -- 9th February. The "d" release was a test only, and went to specific folks only.

 

Pete

[Oliver Grützmann]

Hi Pete,

 

thanks alot. In that case I'm sure I had the .949c version.

 

It's more about testing, if I know that it's coming with FSUIPC I can give LM something to start with. Else I know I need to look elsewhere.

For me it's the easiest to roll back the change I made before the problem occured. If it's still there, at least I have one thing ruled out.

 

I'll send you my email via PM.

 

Oliver.

[jabloomf1230]

Exception code 0xc0000374 (heap corruption):

 

https://social.technet.microsoft.com/Forums/windows/en-US/578396fd-29c0-4dd3-b002-5e2e8df1b398/exception-code-0xc0000374?forum=w7itprogeneral

 

http://blogs.msdn.com/b/calvin_hsia/archive/2015/01/30/10590143.aspx

 

I saw this exception previously, but I haven't seen it in probably 6 months. At the time, it had something to do with MyTraffic 5.4, but the exact cause was not determined.

[Oliver Grützmann]

Exception code 0xc0000374 (heap corruption):

 

https://social.technet.microsoft.com/Forums/windows/en-US/578396fd-29c0-4dd3-b002-5e2e8df1b398/exception-code-0xc0000374?forum=w7itprogeneral

 

http://blogs.msdn.com/b/calvin_hsia/archive/2015/01/30/10590143.aspx

 

I saw this exception previously, but I haven't seen it in probably 6 months. At the time, it had something to do with MyTraffic 5.4, but the exact cause was not determined.

Hi,

 

found Mytraffic as well, but I don't use it. What I also found was the term "stackhash" in the error log, which, it seems, can be caused by almost anything.

Maybe it's really just coincidence that this started after installing a new FSUIPC Version.

 

I'm currently doing some testing, but (un)fortunately unable to reproduce for now (2 hrs of flight).

 

Worst thing is being absolutely clueless.

[Oliver Grützmann]

Hi,

 

me again. Looks like I might have sorted the error out. Seems like it was one corrupt AI model which never appeared before. Must have hit this model (or at least a similar one, which vPilot matched to this model) in two subsequent flights just after installing the new version of FSUIPC. I was able to reproduce the exact same error by running the model matching check of vPilot.

 

Looks solved. Thanks for your help here.

 

Oliver.

[Pete Dowson]

Looks solved. Thanks for your help here.

 

Okay. Thanks for clearing it up here.

 

Pete