Bluetooth Passthrough = "ScheduleEvent from wrong thread (non-CPU)"

[deadkat37]

Hi,

I am having a problem with bluetooth passthrough causing dolphin to throw "ScheduleEvent from wrong thread (non-CPU)" errors  

The error is intermittent (80% of all startup attempts throw the error before any output to the screen) but "House of the Dead 2+3 Return" reliably throws the error at the menu after selecting which game to play.

Here is all the technical info that might be useful:

• Windows 10 Home x64 / i7 6700T / 8Gb RAM / NVIDIA GeForce 960
  
• Dolphin development build (<7 days old) / running fullscreen with -batch -exec
  
• Bluetooth adaptor is an AZIO BTD-V201 with drivers installed via Zadig as per dolphin-wiki instructions (no issues or surprises during install)
  
• Bluetooth Passthrough is configured as FORCE in the dolphin.ini because of internal bluetooth and a Mayflash Dolphin Bar which were being selected ahead of the AZIO (I have also tried the AZIO adaptor in a couple of different USB ports)
  
• Official Nintendo Wiimote (original and motionplus)
  

For the moment I have had to fall back to Emulated Bluetooth via the Dolphin Bar (which is a shame because the 20% of successful startups really sold me on Passthrough!)

Any assistance would be greatly appreciated...

[deadkat37]

When the "ScheduleEvent from wrong thread (non-CPU)" issue is triggered, selecting "Continue = Yes" always seems to continue emulation without any problems.

Could a default be added to the dolphin.ini in conjunction with the disabled Panic Handlers option to suppress the error pop-up and assume continue?

...does anyone have any idea what the underlying problem is that causes the error? I have tried disabling Dual Core because it sounds CPU related, but I am blundering around in the dark.

[leolam]

That sounds like a bug. Do you know how to use a debugger? If so, could you put a breakpoint on the assert call and figure out what is triggering it by looking at the backtrace? We can't do much without having more information, given that we can't reproduce it.

[deadkat37]

That sounds like a bug. Do you know how to use a debugger? If so, could you put a breakpoint on the assert call and figure out what is triggering it by looking at the backtrace? We can't do much without having more information, given that we can't reproduce it.

many thanks for the reply - I'm afraid I don't have a development environment to put in a breakpoint (and my development skills are 25 years out of date...)

could you provide me with a debug build that dumps the info you need?

thanks again

[trihy]

You could at least try older builds to see when the problem was introduced.

[leolam]

many thanks for the reply - I'm afraid I don't have a development environment to put in a breakpoint (and my development skills are 25 years out of date...)

could you provide me with a debug build that dumps the info you need?

thanks again

Sorry, but I wouldn't be able to give precise steps (as I don't really use Windows). I think you may need to use Visual Studio? Honestly, it's a bit of a hassle, so if you can't put a breakpoint, I'd still suggest opening an issue on the bug tracker.

[JosJuice]

We've released the new version 5.0-5620 that will show the event name when you get the "event scheduled from wrong thread" error. Please download it and tell us what the error message says. This won't give us all the details we would get from a stack trace, but it'll at least give us one detail.

[deadkat37]

We've released the new version 5.0-5620 that will show the event name when you get the "event scheduled from wrong thread" error. Please download it and tell us what the error message says. This won't give us all the details we would get from a stack trace, but it'll at least give us one detail.

many thanks - downloading it now.

unfortunately my messing around with the bluetooth adaptors resulted in my bluetooth being completely. System restore did not fix it, so I have had to revert to a 2 month old HDD image and rebuild my PC back to it's former glory.

no idea if the new 'old' PC build will exhibit any of the same symptoms - but I will give it a go and let you know.

[deadkat37]

no idea if the new 'old' PC build will exhibit any of the same symptoms - but I will give it a go and let you know.

success (sort of...) still getting the problem and the new build reports the following extra info:

A "IPCEvent" event was scheduled from the wrong thread (non-CPU)

[deadkat37]

We've released the new version 5.0-5620 that will show the event name when you get the "event scheduled from wrong thread" error. Please download it and tell us what the error message says. This won't give us all the details we would get from a stack trace, but it'll at least give us one detail.

Hi,

the new build reports the following extra info:

A "IPCEvent" event was scheduled from the wrong thread (non-CPU)

As per trihy's reply, I am going to try older builds (starting at 5.0-4869) to see if the problem existed back then.