P3D v4.5 (or 4.4) Crash with AIseparation or RC4

[Quoth]

31 minutes ago, Pete Dowson said:

Yes, It certainly seems so. It's related to trying to control AI in Slew mode via FSUIPC offsets 2900 & 2904.

Pete

 

Thanks Pete, enjoy your holiday!

G

[Quoth]

Hi John

I get the exactly the same crashes as Ray only mine occurs after approximately one hours flying time. I use egkk as my start point... 

Regards

G

[Ray Proudfoot]

Hi John,

Use a default jet. The Raptor should suffice unless you have something else you prefer to use.

Position at stand 414 EGLL. Load the plan I provided into RC4 - EGLL-EHAM. In RC4 on the Controller page enter your gate number (414). Close that dialogue box. Open General and select an airline. I use British Airways PLC. Enter a flight number as you will be referred to by that. 123 is fine. Your call-sign is now Speedbird 123. In RC4 open the keyboard settings. Note what key actions are required for the various actions.

Go into P3D weather and set whatever you want but make sure winds are from the west. Close the weather dialogue.

Start RC4. A SimConnect window should open on the P3D display. Using the relevant keys tune to ATIS (optional) then CLEARANCE. You will be given an initial altitude and squawk code. Unless you want to fly the route just ignore those values. It's probably 1 or 2 for ATIS and CLEARANCE. Use 7 for acknowledge. ATIS loops until you change frequency.

Slew backwards from 414 to give you space to taxi. Come out of slew mode and set parking brake. Start engines. When you're ready to taxi contact GROUND. You will be instructed to taxi to Runway xxx. It might be 27L but if not then acknowledge and then press the option for runway selection. Choose the one for 27L.

Start taxiing to 27L. It's only a short taxi. Make sure you turn to the right and head towards the end of the runway, that's important.

Half-way there you'll be told to contact or monitor TOWER. Change to TOWER frequency. Contact Tower means you're number 1 for departure. Monitor means there's another aircraft ahead of you probably the other side of 27L coming from Terminal 2/3.

Hold short until you're cleared to Line Up and Wait. Taxi onto 27L. As you align with the runway heading P3D may crash there and then. It's happened to me a few times. If not, good. You'll be given take-off clearance. Acknowledge and throttle forward. As soon as your wheels leave the ground P3D should initially freeze for a few seconds and then CTD.

I can't recall being able to depart EGLL with 100% AIG traffic with RC4 running. Without it it's not a problem. Good luck!

[Drumcode]

I'd hate to point out the obvious, but a large amount of AI is needed in these tests. John, not sure whether you have all of the AIGAIM packages installed, I don't think you'll be able to reproduce it running UTL as it is not injecting enough traffic into P3D.

[Ray Proudfoot]

@Drumcode, don't worry, he has 300+ Ai with the AIG package installed. 😉

[robschofield]

Pete/John - that is the same fsais version (1.1) I am using.

I will load up 1.5.1 on Saturday and EGGL and see what happens when slewing gets underway and report back.

 

Regards

 

Rob

[Ray Proudfoot]

On 6/5/2019 at 12:15 PM, Quoth said:

I get the exactly the same crashes as Ray only mine occurs after approximately one hours flying time. I use egkk as my start point... 

I cannot recall any crashes of this type so far into a flight. Would you try the EGLL test please? If you get the .NET crash followed by two FSUIPC crashes within a couple of seconds of the wheels leaving the ground that replicates mine and is reproducible on a third system which adds weight to the argument.

[Drumcode]

6 hours ago, Ray Proudfoot said:

I cannot recall any crashes of this type so far into a flight. Would you try the EGLL test please? If you get the .NET crash followed by two FSUIPC crashes within a couple of seconds of the wheels leaving the ground that replicates mine and is reproducible on a third system which adds weight to the argument.

I get them on roll, shortly after take off, shortly into the flight (maybe 40NM from the origin), on descent, on final, they are all over the place, but the event viewer does show .NET followed by one more or two more errors one of them being related to FSUIPC. 

[Ray Proudfoot]

4 hours ago, Drumcode said:

I get them on roll, shortly after take off, shortly into the flight (maybe 40NM from the origin), on descent, on final, they are all over the place, but the event viewer does show .NET followed by one more or two more errors one of them being related to FSUIPC. 

That's helpful but the reason I'm asking you to perform my test is because it's at a specific place. That is important when trying to replicate a problem.

[Drumcode]

Could we please get an update on this? Is it after all LM's issue? I apologize for being impatient, I just would like to know whether there's any light at the end of the tunnel. I would hate having to give up using RC for another ATC option, and I love my AI. 😄

[Ray Proudfoot]

@Drumcode, Pete is on holiday now and I know John has been working on finding a solution so please be patient. You will hear from John in due course.

The short-term solution is to reduce Ai to around 60%. Not ideal but it should prevent the crashes for now.

[John Dowson]

As Ray says, this is going to take some time, especially with Pete being away. Its not an LM issue,  it seems to be a timing issue in FSUIPC and I can't replicate in a debug version, so its tricky to track down. Just reduce AI traffic for the time being.

I will post when I find more info.

John

[Quoth]

Hi

Flew the same scenario as Ray, from stock egll to stock eham, no add ons except Aerosoft 319, AIM OCI (100%) and RC4. Crashed on approach to eham, approximately 40mins flying time. No problems taxiing, or take off.

Hope that’s of some help

Regards

Q

[Mark_Fox]

On 6/4/2019 at 1:22 AM, robschofield said:

Did testing last night having found a copy of v 1.5 in another post on this topic.

 

Outcome - FSAIS works fine in either P3Dv4.4 or P3Dv4.5 with FSUIPC v5.1.5, but as soon as you use 5.1.5.1 or 5.1.5.1a or b then irrespective of whether you are using P3dv4.4 or 4.5, it will crash when ai slewing becomes active. I think John has picked this up in the other post on this topic, has got a copy of RC4, and suspect whatever fix does the job for RC4 will similarly work on any other application which makes use of FSUIPC as its interface into the simulator.

I can confirm what robscofield is seeing. I had tracked down a crashing issue to AI Separation. I would received this error:

Quote

Prepar3D.exe    Stopped working    5/17/2019 4:41:40 PM    All Users    0xc0020001    The string binding is invalid. 0x00039129    KERNELBASE.dll    10.0.17763.475    P:\Prepar3D v4\Prepar3D.exe   

 Taking Rob's advice, I went back to 5.15 from 5.151. AI Separation may not always work due to what I believe is an issue with user position within AI Sep itself, but it no longer crashes P3D V4.512 to the desktop on 5.15.

I have attached a text file with the error as last seen in AppCrashView. Hope this may help in some way.

Respectfully,

Mark Fox

[UniformAlpha2014]

Hi Guys

I had that same crash with AIM OCI and FSUIPC the last days as well. So I tested around a little bit. 

For me, I found out, when going back to FSUIPC version 5.11, it works fine with the AIM OCI. The only downside is, I can't see the complete runway assignment of approaching AI aircraft in AISeparation. It is only somthing like: "EHAM 18", when it should be "EHAM 18R". 

[Drumcode]

Is there a way to download version 5.11 anymore? I'd like to check whether my Radar Contact issue will also be solved by running an older version.

[John Dowson]

Do you mean 5.15?    @UniformAlpha2014 reports 5.11 being ok, but I assumed that was a typo. I can post 5.15 if you really need it (i.e. reducing AI traffic is not an option), but this is not supported, so please don't post any issues if you use it. Let me know.

I'm pretty certain that the problem is due to heap corruption somewhere, but its a bugger to track down. The release (commit) where the problem first arises is not related to where the problem was introduced. Any heap analysis tool I've tried crashes in P3D before it even touches the FSUIPC code, which is also worrying (and am working on to raise with LM).

Thanks for all the info so far, but no point in posting more really - just adding noise to the topic.

Cheers,

John

[UniformAlpha2014]

no it really is version 5.11 😉 

found it on a old HDD.

[John Dowson]

Does that mean 5.15 doesn't work for you?

[UniformAlpha2014]

I have tested the version in my simmarket account and the one's you posted on this other post some time ago. They aren't working for me 😕 

[Drumcode]

Anyone who can share 5.11 with me gets a big hug.

[John Dowson]

The version on SimMarket is 5.151, and the ones posted since on other posts have been later versions - 5.151a, & 5.151b. Could @Drumcodeand @UniformAlpha2014 therefore please try 5.15..

5.15 is attached - rename to FSUIPC5.dll once downloaded.

FSUIPC5.txt

Edited June 14, 2019 by John Dowson
Attachment added

[Drumcode]

I will give it a go this evening and report back. I also have an old version of 5.1.0 and will fire it up if 5.15 fails with traffic at 100%.

[Drumcode]

Reporting back... Completed two evening flights DEN-SLC-DEN and no CTD. Not sure I'm completely out of the woods yet, but tomorrow I'll try DEN-IAH, two big AI hubs. 

[Quoth]

52 minutes ago, Drumcode said:

Reporting back... Completed two evening flights DEN-SLC-DEN and no CTD. Not sure I'm completely out of the woods yet, but tomorrow I'll try DEN-IAH, two big AI hubs. 

Which version was that with?