Adding FSUIPC5 to DLL.XML causes P3D v4 to crash

[Ray Proudfoot]

Hi Pete,

Thought I'd help by posting here. Following my earlier post on AvSim I was able to get p3D v4 to show the free flight menu with no EXE.XML or DLL.XML files in my appdata\roaming\L-M\p3D folder. Five reboots and five successful loads.

I reinstalled FSUIPC5 so it would place an entry in DLL.XML and rebooted. Checked the entry before starting P3D and it looked fine. Launched P3D and it crashed before free flight window. But the only difference from before was the inclusion of FSUIPC5.

I'm attaching various files which I hope will help. As you can see the latest fsuipc5.log didn't show all the hardware but in the JoyScan.csv you can see they're all present. I'm also including a previous successful fsuipc5.log for confirmation my hardware looks okay.

Hope you can spot something because it's driving me crazy.

FSUIPC5.ini

FSUIPC5.log

FSUIPC5.JoyScan.csv

FSUIPC5_prev.log

DLL.XML

[Pete Dowson]

3 hours ago, Ray Proudfoot said:

As you can see the latest fsuipc5.log didn't show all the hardware but in the JoyScan.csv you can see they're all present.

They are all in the INI file:

[JoyNames]
AutoAssignLetters=No
T=Saitek Pro Flight Throttle Quadrant (USB)
T.GUID={D78B8DC0-E4DE-11E8-8012-444553540000}
0=Saitek Pro Flight Throttle Quadrant (USB)
0.GUID={D78B8DC0-E4DE-11E8-8012-444553540000}

Y=Saitek Pro Flight Yoke
Y.GUID={D738B2D0-E4DE-11E8-800B-444553540000}
1=Saitek Pro Flight Yoke
1.GUID={D738B2D0-E4DE-11E8-800B-444553540000}

U=Saitek Pro Flight Throttle Quadrant (USB)
U.GUID={D78B66B0-E4DE-11E8-8010-444553540000}
2=Saitek Pro Flight Throttle Quadrant (USB)
2.GUID={D78B66B0-E4DE-11E8-8010-444553540000}

P=CH PRO PEDALS USB
P.GUID={D7388BC0-E4DE-11E8-800A-444553540000}
3=CH PRO PEDALS USB
3.GUID={D7388BC0-E4DE-11E8-800A-444553540000}

Or is that your doing, editing?

The log file is stopping, as before, in midst joystick scan. Very suspicious. See what happens if you unplug them all first. Then try one at a time.

You said on the AVSIM Forum that it booted okay second and subsequent times. This suggestes something odd about initialisation of the devices causing DirectInput to crash.  If you can nail it down to one device, see if waggling its axes or pressing its buttons before loading P3D makes a difference.

You could also get more detals about the scanning by extra logging. Add these to the [General] section in FSUIPC5.INI:

Debug=Please
LogExtras=x200000

Do you have P3D4.3 in stalled o this new PC yet? If not and you plan to I'd do it now and test on that too. Because it might be a failing USB port (you could try moving them aound too, of course, to check that). Bear in mind that USB ports are paired -- i.e. same hardware chippery for pairs.

Pete

 

 

[Ray Proudfoot]

Hi Pete,

Time for bed but I’ll look at removing flight sim devices and adding them back one at a time. The Saitek yoke acts as a hub for the two throttle quadrants. Only the yoke and the CH Pedals are plugged directly into USB ports on the PC.

Earlier in the evening there was an entry in FSUIPC.ini saying *Missing Joystick*. I checked the button entries in the ini and I had pasted some hat switch views from the forum into it without changing the letter referencing the device. Once I removed those the *missing joystick* entry disappeared.

Those entries were added by me in line with the guidance in the user manual.

I also saw a message about DirectInput or words to that effect when P3D was starting but it was temporarily hidden by the splash screen. That relates to DirectX if I remember right so it could be related.

I later ran the utility to check DirectX and it completed okay. Rebooted and ran P3D but it crashed.

I’ll try renaming the ini first, then unplugging USB devices and then add extra logging. Hopefully one of those will either solve it or give us a clue as to what’s going on.

The most confusing part is it running okay on a second attempt.

Don’t understand your last question. Do you mean P3D v3.4? If so, answer is no, not yet.

[Pete Dowson]

8 hours ago, Ray Proudfoot said:

Don’t understand your last question. Do you mean P3D v3.4?

Yes, sorry -- typo.

I was thinking that since you've been using 3.4 a while on your old PC wirhout these crashes, that checking it out on the new one would check whether it had a similar problem -- so pointing to something more hardware related than software (or maybe Windows of course!).

One other thing: as well as the log with the extra logging (on a run which crashed, not a good run), the Joyscan csv file from that run would be useful. It would also show what stage in the FSUIPC scanning it got to. The file you supplied erlier shows a completed scan, not one crashing part way as the log does (there's no "prev" CSV, so you'd need to get that file before your usual re-launching.

Also, I assume you have P3D4 set with controllers disabled. If so, then P3D4 is not scanning devices either. (Actually, I'm not sure whether it does that during loading or not in any case). P3D has two scanning modes -- normal DirectInput, like FSUIPC, and "direct" which I think means actual USB serial port interface use, much like FSUIPC's COM HID library facilities in Lua. So, there are all these variations to test, too, to try to pinpoint where the problem may lie.

Pete

 

[Ray Proudfoot]

Hi Pete,

Just a quick one as I have to help a friend with Windows 10 this morning.

P3D v3.4 was fine on the old PC but of course that was with W7 and 10 is a very different animal as we know. It shouldn't take long to get v3.4 installed but I'll try what you suggested last night first as those are fairly quick to test and report.

Yes, the first addon I installed was FSUIPC5 so P3D controllers are disabled and everything is handled through FSUIPC5. It might be useful to disable FSUIPC5 (rename DLL.XML) and enable controllers in P3D v4.3. That would force it to scan for my USB hardware. But there'd be no log of course so if it failed it wouldn't point to what was responsible.

I'll work on this early afternoon.

[Ray Proudfoot]

Hi Pete,

The first thing I tried was to rename fsuipc5.ini forcing FSUIPC5 to generate a new one. P3D v4.3 crashed and attached are the files created. I'll reboot and add the extra checking into my original fsuipc5.ini.

The Event Viewer log was as follows:-

Faulting application name: Prepar3D.exe, version: 4.3.29.25520, time stamp: 0x5b2c3263
Faulting module name: ntdll.dll, version: 10.0.17763.1, time stamp: 0xa369e897
Exception code: 0xc0000374
Fault offset: 0x00000000000fb349
Faulting process ID: 0x2430
Faulting application start time: 0x01d48593bf50a20e
Faulting application path: D:\P3Dv4\Prepar3D.exe
Faulting module path: C:\WINDOWS\SYSTEM32\ntdll.dll
Report ID: 8043d4d5-df6b-44fa-8d88-f155c0a3240e
Faulting package full name:
Faulting package-relative application ID:

FSUIPC5.log

FSUIPC5.ini

[Ray Proudfoot]

Hi again Pete,

I rebooted and added extra logging option to my original fsuipc5.ini. P3D still crashed. Various files attached. The crash report was virtually identical to above but with different fault offset.

Faulting application name: Prepar3D.exe, version: 4.3.29.25520, time stamp: 0x5b2c3263
Faulting module name: ntdll.dll, version: 10.0.17763.1, time stamp: 0xa369e897
Exception code: 0xc0000374
Fault offset: 0x00000000000fb349
Faulting process ID: 0x3004
Faulting application start time: 0x01d48594cb177beb
Faulting application path: D:\P3Dv4\Prepar3D.exe
Faulting module path: C:\WINDOWS\SYSTEM32\ntdll.dll
Report ID: f326fd3c-feb0-4cd3-b43f-abfa2487ef00
Faulting package full name:
Faulting package-relative application ID:

I shall now remove yoke, pedals and 2 x throttle quadrants, reboot and try again.

FSUIPC5.JoyScan.csv

FSUIPC5.log

FSUIPC5.ini

[Ray Proudfoot]

Hi Pete,

Shut down PC, removed yoke, pedals and throttles. Restarted and launched P3D v4.3. It showed the free flight menu and I went with the default. After 10-15 secs I exited P3D. Files attached.

I shall now shutdown, add pedals only and try the sim again.

FSUIPC5.ini

FSUIPC5.log

FSUIPC5.JoyScan.csv

[Ray Proudfoot]

Hi again Pete,

I plugged just the CH Pro Pedals into a USB2 port - they aren't recognised in a USB3 one - and P3D crashed. Just with one device. Files attached.

I'm wondering if this is anything to do with Windows update 1809. And why just on first boot and no thereafter? I'll unplug pedals and try yoke on its own.

FSUIPC5.log

FSUIPC5.ini

[Ray Proudfoot]

Final test. Just the yoke plugged into a USB2 port and P3D crashed before the free flight screen. Log attached.

I think there's a problem with my USB ports. Software rather than hardware. I'll try the client uninstall / reinstall option. Last chance saloon. More later.

FSUIPC5.log

[Ray Proudfoot]

I plugged yoke, throttles and pedals back in and rebooted. Started P3D but rather than crash it just hung with the splash screen showing. Final logs attached. Will launch it again and let's see what happens...

And it launches successfully. Now going to go for client update...

FSUIPC5.log

FSUIPC5.JoyScan.csv

[Pete Dowson]

51 minutes ago, Ray Proudfoot said:

The first thing I tried was to rename fsuipc5.ini forcing FSUIPC5 to generate a new one. P3D v4.3 crashed and attached are the files created. I'll reboot and add the extra checking into my original fsuipc5.ini.

The Event Viewer log was as follows:

That sort of Windows crash data would probably be of more use to L-M than me,

35 minutes ago, Ray Proudfoot said:

Shut down PC, removed yoke, pedals and throttles. Restarted and launched P3D v4.3. It showed the free flight menu and I went with the default. After 10-15 secs I exited P3D.

So with no devices scanned it is okay!

11 minutes ago, Ray Proudfoot said:

Final test. Just the yoke plugged into a USB2 port and P3D crashed before the free flight screen. Log attached.

I think there's a problem with my USB ports. Software rather than hardware. I'll try the client uninstall / reinstall option. Last chance saloon. More later.

That shouldn't be the final test! There are more to narrow it down further. As I suggested earlier,  I think you should test with:

1. No FSUIPC, devices plugged in, controllers enabled in P3D.

2. Same but using the other P3D mode (i.e changed the "direct" option)

3. Using completely different USB port(s).

Going back a few emails, trying to find logs which sohwed things, you said

50 minutes ago, Ray Proudfoot said:

I rebooted and added extra logging option to my original fsuipc5.ini. P3D still crashed. Various files attached.

but the Log file you attached showed a good session with a normal exit!!

Confused! :-(

Pete

 

[Ray Proudfoot]

Final logs from successful launch in case they're helpful.

FSUIPC5.ini

FSUIPC5.log

FSUIPC5.JoyScan.csv

[Pete Dowson]

3 minutes ago, Ray Proudfoot said:

I plugged yoke, throttles and pedals back in and rebooted. Started P3D but rather than crash it just hung with the splash screen showing. Final logs attached.

The log shows all the devices scan okay. The hange was after it loaded the default flight.

Pete

 

[Pete Dowson]

3 minutes ago, Ray Proudfoot said:

Final logs from successful launch in case they're helpful.

Not really. All the logs for success show success. The logs for failure all show a hang in the scanning except the odd one where it was later, after all the scanning was okay. I'm not sure now what you are changing between each in any case.

Pete

 

[Ray Proudfoot]

2 minutes ago, Pete Dowson said:

So with no devices scanned it is okay!

That shouldn't be the final test! There are more to narrow it down further. As I suggested earlier,  I think you should test with:

1. No FSUIPC, devices plugged in, controllers enabled in P3D.

2. Same but using the other P3D mode (i.e changed the "direct" option)

3. Using completely different USB port(s).

Going back a few emails, trying to find logs which sohwed things, you said

but the Log file you attached showed a good session with a normal exit!!

Confused! :-(

Pete

 

Hi Pete,

Okay, I'll disable DLL.XML and enable controllers in P3D. Do you know the cfg entry for direct mode?

I have only 4 USB2 ports so I'm limited. GoFlight stuff, mouse / kbrd take two.

Check logs again. I'm sure I posted correct ones for each test.

[Pete Dowson]

Good past on the L-M forum -- succinct and to the point.

Pete

 

[Ray Proudfoot]

Hi Pete,

No reply to the L-M post but I see 4.4 is released tomorrow. Rather than reinstall 4.3 I might as well install 4.4 and see if that changes anything.

[Pete Dowson]

2 hours ago, Ray Proudfoot said:

No reply to the L-M post

Yet, you mean ...  even I don't get immediate responses! 😉

Pete

[Ray Proudfoot]

Maybe I'm impatient but I would have expected a question or request for additional info.

You missed the e off your name. Shall I emulate Geordies and call you Pet from now on? 😁

[Pete Dowson]

41 minutes ago, Ray Proudfoot said:

Maybe I'm impatient but I would have expected a question or request for additional info.

You posted it late yesterday. They are VERY busy at present (4.4 is imminent, remember). Even in "slacker" times I don't see immediate responses which it would be -- bear in mind they aren't retired hobbyists and have workng hours (and the US not in Europe).  How many developers give you same day responses let alone within hours as it would have had to be to arrive before now?

They aren't all as daft as me -- I sohuldn't do it really. i spend more time typing (and mis-typing) answers here than I do anything else! 

Pete

 

[Ray Proudfoot]

I didn't know about 4.4 so I'm happy to give them time. You're not daft, it's excellent customer service. 👍