DXGI_ERROR_DEVICE_HUNG error: Is there a solution?

[SteveW 1,792]

5 hours ago, squerble said:

Pete, do you do anything else on the PC running P3D when this happens? I've noticed my PC is much more likely to give me this error if I'm doing something else on the PC at the same time, be this using Chrome or opening a PDF. I've yet to try a "proper" clean with DDU but sometimes it'll crash as soon as I open a PDF - I'm wondering if it's something to do with P3D losing focus, at least on my end.

That's what I'm talking about. Pete should factory default the main page and use a P3D specific profile. The P3D system is basically running on the desktop with all those browser windows. Not sure if it's his problem but to fail to eliminate the possibility in the first place is a mistake.

[glider1 90]

I replaced the PSU for nothing!

New PSU, new GPU (2080).

Passes all the stress tests.

Still a random DEVICE HUNG!

Sh#$%*$%

I think I might have got it once on 4.3 a few months ago,

But on 4.4 it is every second flying session what gives?

[wyman_hd 232]

On 12/4/2018 at 5:19 AM, Pete Dowson said:

Only recently (since P3D4.3, and still with P3D4.4) I started to get these DXGI device has hung messages. Originally, when they started, they only occurred at dusk or at night. But I've just tried re-enabling my FTX England scenery to see what it looks like in version 4.4.

Now the crash occurs in the daytime, 100% consistently, within a couple of minutes of the scenario starting up (my default, at UK2000's EGCC).

The GPUs (I have 3 x 1080Ti's in SLI formation) are never loaded to more than 45% (according to Task Manager), so I don't understand what is going on. Is this a P3D problem, or an nVidia driver problem? The error report appears to be from P3D. I have tried various nVidia drivers, Currently I'm on 416.94.

My installation is pretty well loaded up with addons as well as Orbx, but I tried with the FTC OpenLC Europe disabled, and with all my UTX layers disabled too, to no avail. With no FTX Europe regions enabled all is well -- for daytime flying. I can't use it at night because of the same errors.

I'll ask on the P3D forum too. I need a solution, or my Orbx purchases are useless (I have all of the FTX Europe country regions and was looking forward to trying them, but I can't get passed the first couple of minutes after loading).

Whilst awaiting answers I'll continue experimenting. Things to try:

1. Disabling SLI
2. Changing drivers yet again (go for the latest first ...
3. ... er: there is no 3 ... yet.

Thanks,
Pete

Well, on the AVSIM CTD forum, I'm the one who actually posted about this as well on the CTD forum. 

But, I had a faulty addon, (plus it was p3d v3.4!) and I don't have and will never have the money for SLI.

On 12/4/2018 at 9:29 AM, Pete Dowson said:

I'm seeing it loaded up to 90% in heavy situations

That's happening with my GTX 1070! (I guess it is normal!)

All I can say is that, just use 2 of the cards, and if framerate is over 30, then don't even bother. 

You can't really tell the difference between 30 FPS, and 165 FPS, with your bare eye, so if it's more than 30 FPS, you should be happy with that! (Especially in dense situations!)

Edited December 11, 2018 by arwasairl

[SteveW 1,792]

"With no FTX Europe regions enabled all is well -- for daytime flying. "

From Pete's posts it seems there are situations from within his scenery setup that alter the error behaviour, supposedly with FTX disabled daytime flying is still possible without error.

This reflects the idea I put forward to test with reduced architecture by lowering the settings. not so to be 'reduced' - what I'm attempting here is to try to eliminate certain buildings and features, rather than reduce the detail levels. (The differences between day and night brings a different set of features).

So from ensuring the profiles are not interfering, and that the set of architecture has been altered, the next step to test would have to be to ensure P3D is stock, that means stock.

 

Edited December 11, 2018 by SteveW

[Pete Dowson 558]

On 12/10/2018 at 9:05 AM, squerble said:

Pete, do you do anything else on the PC running P3D when this happens?

No. Other processes are running, of course, but the focus is on the P3D display across the three projectors.

19 hours ago, SteveW said:

Pete should factory default the main page and use a P3D specific profile.

In NCP you mean? Why create a profile at all? I've never used one and have only been getting these DXGI errors recently (with 4.4 really -- just once before 4.4, since installing P3D4 on this system 18 months ago!).

19 hours ago, SteveW said:

The P3D system is basically running on the desktop with all those browser windows.

No browser use at all on the P3D4 PC. It is dedicated to the cockpit only. I cannot easily use other programs on the curved 210 FOV screen -- most of the time, with my RP narrow vision, I can't even find the mouse pointer. To set things up (whether NCP or other) I have to use TeamViewer from another PC.

5 hours ago, arwasairl said:

You can't really tell the difference between 30 FPS, and 165 FPS, with your bare eye, so if it's more than 30 FPS, you should be happy with that! (Especially in dense situations!)

In dense situations I am doing well if it reaches 20 fps! I'd be over the moon if it never dropped below 30 (which is effectively my limit in any case, via VSync and Projectors set to 30Hz)!

Pete

 

[SteveW 1,792]

3 hours ago, Pete Dowson said:

 

In NCP you mean? Why create a profile at all? I've never used one and have only been getting these DXGI errors recently (with 4.4 really -- just once before 4.4, since installing P3D4 on this system 18 months ago!).

 

For the purpose of a test Pete that's all I'm saying - and not just to you.

I'm not suggesting anyone change their setups or anything like that - just to eliminate possibilities.

[SteveW 1,792]

3 hours ago, Pete Dowson said:

No browser use at all on the P3D4 PC. It is dedicated to the cockpit only.

 

 

Maybe so but Windows is a dense browser enabled system under the hood, which is not removable. So I'm not saying you might have a browser open at all, but the functionality is active.

[SteveW 1,792]

...I mean Pete - I've just been trying to come up with some tests you haven't done - so you discuss them and it takes longer than trying them out. Even so I suspect some part of the scenery is creating a confluence of events that confuse the card now. Things move on in P3D and it's possible some loose end or loophole been tied up that may have enabled something to get away with it in the past.

If you can find the cause in the stock system so that LM can re-create it you can present it to LM.

Then put your system back exactly how you like. 

[SteveW 1,792]

...by scenery I mean even planes and other objects too. Could be an AI or static plane or vehicle rather than an airport.

[Bob Scott 7,750]

Are people seeing this error also using shader mods like PTA, EnvShade etc?  I wonder if a bug in the add-on shader code could be causing this, or contributing to it.

 

[SteveW 1,792]

1 hour ago, w6kd said:

Are people seeing this error also using shader mods like PTA, EnvShade etc?  I wonder if a bug in the add-on shader code could be causing this, or contributing to it.

 

Anything related to shader mods should be eliminated in the first place as well. The system has to have the possible variables removed and anything unrelated should be eliminated, reducing the system to the most basic set that can invoke the problem.

Someone noticed that using browsers could have an impact on the behaviour. Given that the system is built on COM and that the desktop window is an explorer window, like a browser window all use the same stuff. So the browser interaction can be a clue - doesn't need a browser working on the desktop as Pete puts it.

When the Dx11 manual puts it plainly as I said earlier, the DXGI is partial to incorrect data formats and will hang. Temporary Debug Recovery Mode zero should not be used in case of more extreme errors.

 

 

[planenut 19]

Prior to updating to v4.4 I only got the DXGI error only occasionaly but not recently.  After upgraded to 4.4 I also installed the latest driver at the time (417.01 I think).  After that I started getting the error pretty regularity.  The last time I got it I had just loaded P3d with my default plane (LM F16) and had the error and crash within about 3 minutes without doing anything.  Plane was just sitting on the ramp at KSEA.  I don't fly this plane - just use it as my default.  So I reverted back to an older driver (391.35) and haven't had any problem since.  Don't know if this a fix but it sure seems to work for me (knock on wood).  Hope this may help someone in their troubleshooting this aggravation.

Jim

My computer specs are:

I7 6800K @ 3.40 GHz,  16 GB Ram, GTX 1080

[pgde 571]

Here is an update to my 1080ti saga. The 1080ti is on its way back to Gigabyte due to this post's subject error. I stopped by Best Buy today and picked up a "test" 2080 to see if I could reproduce the errors. First of all, I can't believe how big the 2080 is. Good thing I have a full size case :) I installed it and have been testing various software applications that were crashing with the original card. Surprise, all apps ran perfectly, even 3D Mark which croaked on the 1080ti. And I also learned that for max performance you need Windows 10. I am on 7 but found a suitable driver on the MSI website (that is what BB had in stock) for Win7. So, IMHO, this locks down the diagnosis that the card is bad. It will be interesting to see what the results of the RMA process with Gigabyte is. And when I get the 1080ti back whether the problem will reoccur. Thanks for everybody's help in this.

Happy Holidays!

P.

[Pete Dowson 558]

9 hours ago, SteveW said:

So I'm not saying you might have a browser open at all, but the functionality is active.

So do you mean that has to be stopped somehow? Some dep -rooted Windows services?

9 hours ago, SteveW said:

Even so I suspect some part of the scenery is creating a confluence of events that confuse the card now. Things move on in P3D and it's possible some loose end or loophole been tied up that may have enabled something to get away with it in the past.

Maybe so, but I been testing all day with no recurrence.

9 hours ago, SteveW said:

If you can find the cause in the stock system so that LM can re-create it you can present it to LM.

That would be neat, but it has only ever occurred he in a particular scenery -- but one which I've been using in P3D4, rigorously (after all, it is EGLL) since first installing P3D4 (and before, in previouus scenery versions).

Before P3D4.4 is occurred just once, a month or two ago. With P3D 4.4. it just started occurring reproducibly -- but now, not since i installed newer nVidia drivers.

1 hour ago, pgde said:

So, IMHO, this locks down the diagnosis that the card is bad.

Or that something in the Video Driver actions for the 1080 are liable to reveal the error, whilst those for the 2080 are not. Also, of course, you have to consider that you are probably driving the 2080 to a lot less than its performance capacity compared to the 1080.

I thought I had it narrowed down to a specific 1080 card (one out of three) but after removing that I got it on one of the others.

Pete

 

[glider1 90]

These things did not work for me on a stock 2080 latest build of W10.

• reducing GPU power limit to 75% in Gigabyte aorus  (DID NOT WORK)
• reducing GPU overclock to default in aorus (DID NOT WORK)
• Cleaning out the driver in DDU in safe mode and reinstalling (DID NOT WORK)

The only lead so far is that the DEVICE HUNG happened when I had LITTLENAVMAP in the foreground maximised with P3D window maximised in the background.

I don' think it matters what monitor the other app is on, it is just whether it has the focus or not. I think it might trigger a DEVICE HUNG at a random time interval while the other app has the foreground it could be milliseconds to minutes before it happens.

The only lead for me so far. Unlikely to be hardware since I replaced the PSU with a new more powerful one and the GPU was stock clocked and throttled back for the test today.

Edited December 12, 2018 by glider1