[BUG FIXED] RTT Remote - Freezing
-
You either have to build it on your own from the source on git or
That’s my version build from the source. It is without installer so unzip and run the exe with administrative rights. No need to deinstall the older version
https://drive.google.com/file/d/177VhX_kTKOSL9vjobD7fzHGk2GupigrP/view?usp=sharing -
Guys, let me try to stop all rumors
There were 2 issues as part of this RTT problem:
1. RTT export itself would freeze
2. BMS itself would freeze and potentially CTDThe BMS freeze was a problem of the RTT export code causing a sync problem with the GFX engine. In short details - The DX11 device context isn’t thread safe in general (i.e you can’t use the immediate context object from other threads), and in this case while the engine was executing draw commands on other threads, the RTT code was using it without the necessary safety. That caused a corruption of the context which eventually led in many cases to a freeze or a crash.
FWIW, I tried to reproduce the problem on my system a couple of times but couldn’t, so it must be luck or randomness or whatever, but for sure the problem is valid and hence was fixed (Both issues needed a fix).
Hello I-Hawk,
I am tired of all the rumors posted here.
Do you have any home remedy for this issue until a patch is released.
I really like the concept of RTT but it freezes all the time after a few minutes.
Please help.Thanks,
-
Until the update is released the only remedy is to disable RTT export in the Falcon BMS.cfg file by changing the 1 to a 0.
-
We use 4.34 U4 until 4.35 U1 is released.
Hopefully it won’t be long until U1, I’d even be happy if it comes with just that fix and a few others (like the tanker). -
I’m experiencing the same. The extracted MDF hangs after a few minutes.
Also from time to time the BMS itself hung. I noticed the use of GPU is very high, 100%. I’m using a nVidia GeForce RTX 2070 and usually is at 50% or 60% of GPU usage.
I include the nVidia information.
MibTrol -
I’m experiencing the same. The extracted MDF hangs after a few minutes.
Also from time to time the BMS itself hung. I noticed the use of GPU is very high, 100%. I’m using a nVidia GeForce RTX 2070 and usually is at 50% or 60% of GPU usage.
I include the nVidia information.
MibTrolIt’s a confirmed bug with 4.35 and nothing you can fix it apart of waiting for U1. Changing your Nvidia settings won’t change anything in regards of the bug
Gesendet von meinem SM-G930F mit Tapatalk
-
I have noticed that when you have your ACMI buffer at 20MB it will quickly freeze your RTT extracted MFDs but if you crank it up to 500MB or 1000MB you get freezes much later. Can others verify this. I went back and forth from 500MB back to 20MB and could get it to freeze quickly with 20MB and can fly a longer time before the freeze with the high 500MB or 1000MB settings. So as a work around set the ACMI Buffer to 1000MB. It is regardless of recording.
-
And when you don’t record one?
-
And when you don’t record one?
I can confirm this. It seems that the ACMI size also increased DX11 buffersize. I changed it to 1000MB and since than no more freezes.
Good find ohommes! Maybe this is a usable workaround till U1 release.
-
I can confirm this. It seems that the ACMI size also increased DX11 buffersize. I changed it to 1000MB and since than no more freezes.
Good find ohommes! Maybe this is a usable workaround till U1 release.
I always record due to -acmi shortcut flag and had my ACMI setting to 1000MB from the very first mission in 4.35 so I can’t repro the described behavior. I can have 5 mission without freeze and then the next 3 mission get a frezze within the first 2 minutes
-
Does, the loitho ACMI Compiler have any effect on freeze cadence?
Just a thought.
Sent from my Phone 2 using Tapatalk
-
Does, the loitho ACMI Compiler have any effect on freeze cadence?
Just a thought.
Sent from my Phone 2 using Tapatalk
Doesn’t make technically any sense as the loitho compiler does nothing while in 3D it just renames the .flt file in order to avoid BMS internal compiler to step in once you leave 3D and then do the compilation
-
Well, 3-4 weeks it’ll be redundant anyway [emoji1696]
Sent from my Phone 2 using Tapatalk
-
I always record due to -acmi shortcut flag and had my ACMI setting to 1000MB from the very first mission in 4.35 so I can’t repro the described behavior. I can have 5 mission without freeze and then the next 3 mission get a frezze within the first 2 minutes
Do you exit and restart BMS between those 5 missions or “in one go” ?
Well, I can’t say what connection ACMI has with RTTex , my acmi is set to 500MB , and by today I never had RTT freeze with “MFD’s for android” ONCE …
-
Where is the option to change the buffer size? CFG file? thanks HC
-
Where is the option to change the buffer size? CFG file? thanks HC
BMS 2D Setup->Simulation Lower right corner “ACMI File Size”
-
Do you exit and restart BMS between those 5 missions or “in one go” ?
Well, I can’t say what connection ACMI has with RTTex , my acmi is set to 500MB , and by today I never had RTT freeze with “MFD’s for android” ONCE …
I’m restarting BMS everytime I come out of 3d regardless if Multiplayer (required you can’t even commit a second time) or single player.
As a developer I have been hunting with 2 other 3rd party devs to get a reliable workaround since the day 4.35 was released and after hours of .dmp debugging all I can say it has to be fixed within BMS code.
According to Bugtracker it’s already done for an upcomming release.All other suggested “potential” workarounds are wild guesses as the Bug is internal related to threat safety so it can be multiply BMS internal threats causing the freeze.
-
The ACMI buffer size isn’t a guess it is reproduce-able observation. And yes it isn’t really a workaround but it does seem related thread handling. Maybe with small buffers you setup a new thread more often for the ACMI (yes this is a guess) but it stands that 500MB or 1000MB does reduce your chances of freeze and this anyone can verify by setting low ACMI buffers. try it and you will get your freeze guaranteed and quickly. So yes large buffer is no guarantee but the probability has gone way done. But I would not fly a MP mission since it still isn’t reliable. So yes we all impatiently wait for U1
-
Can the BMS team Shared with all of us when they think or plan to have the update to correct this issue. At least I don’t fly in 4.35 because I have extracted both MFDs and now it would be very uncomfortable for me to fly in this version.
Thanks in advance
-
I’m still experiencing this behavior. TE randomly freezes only MFDs working with RTT Client. 4.35 last version available today.