Yikes, I’d say that’s a bad! Thanks for keeping us updated on these things Snow. Always appreciate it!

The disarm bug is a very serious issue. I am sure they will get around to fixing it.

Some worrying things with the disarm/failsafe bug in Betaflight 4.3.0 are that:-

1. One of the Betaflight developers has removed the "Bug" tag from the issues suggesting that it's not actually an bug which I think it clearly is.
   
2. Another developer has stated that it probably won't be looked at until the Betaflight 4.4 release which is way too long to wait for such a serious issue IMO.
   
3. The people reporting the issues are now being blamed for running a custom target build, but Dominic Clifton [hydra] who is a respected and experienced Cleanflight/Betaflight developer has clearly shown that the issue is a generic problem which also exists in the official Betaflight source code and so could also occur for any official Betaflight target.
   

It appears to me that so many changes have been made to the source code for the Betaflight 4.3 release that no-one really knows how to rectify the issue without some major rework to some of the code base. I honestly think that this Betaflight release should really have been named Betaflight 5.0 with the huge amount of major changes that have gone into it. I have no idea why they decided to just increment the minor release part of the version number.

They should never have stacked up so many changes in one release.

Fixing these in 4.4 is not a satisfactory way forward.

I do not have access to my quads till towards the end of February, else I would flash my diehard 5” freestyle quad with 4.3 and test/reproduce/report such a dangerous bug.

I can _kind of_ see both sides of this.

The pilot reporting the issue is running their own custom firmware with filter changes that they're not making public.
The Betaflight devs say they can't test it or reproduce it without knowing what was changed.
The pilot says it shouldn't matter, the scheduler should always handle disarms. Even if the load is high.

Honestly, the bug reporter should release their code, but the BF devs are being overly defensive and way too hard-nosed about this.

Some additional comments have appeared in the Betaflight 4.3.0 disarm bug issue from HERE onwards and one of the Betaflight devs has now closed the issue (why?). Bruce Luckcuck [etracer65] has also weighed in as an official Betaflight developer and made a statement that with this disarm bug present, Betaflight 4.3 isn't safe for general use or release until the issue has been fixed...

https://github.com/betaflight/betaflight...1023155655

Bruce Luckcuck [etracer65] Wrote:While it's unfortunate that the original code related to this issue isn't provided, it should be easy to replicate the problem based on the task timings provided above. Clearly the GYRO task is running longer than normal. And the combination of gyro/filter/pid is more than 125us so with an 8K pid loop I would suppose the new scheduler would have problems with that. But the point is that regardless of whatever misbehavior happens in a task, it shouldn't prevent critical functions like being able to disarm. That's the fundamental underlying concern that keeps being ignored. No matter how "delayed" the tasks might get the scheduler should NEVER prevent tasks from running. Continuing to keep sticking your heads in the sand won't solve the problems.

At this point I don't consider Betaflight to be safe for general use and would not recommend it be released until these issues are fundamentally resolved. Otherwise people are going to get hurt and Betaflight's reputation will be permanently tarnished.

(27-Jan-2022, 06:20 PM)SnowLeopardFPV Wrote: Some additional comments have appeared in the Betaflight 4.3.0 disarm bug issue from HERE onwards and one of the Betaflight devs has now closed the issue (why?). Bruce Luckcuck [etracer65] has also weighed in as an official Betaflight developer and made a statement that with this disarm bug present, Betaflight 4.3 isn't safe for general use or release until the issue has been fixed...

https://github.com/betaflight/betaflight...1023155655

Glad to see the adults have arrived to sort things out!

I updated my config to 10.8 and grabbed a toasted AIO FC .. loaded 4.3 onto it..loved the options.. haven’t loaded it to a “functional” quad JUST YET…

I have 4.3.0rc2 loaded on The Leaf and it has performed flawlessly for me so far. In light of the current disarm error concerns, might as well roll it back to 4.2.xx and wait for the next stable release. 10.8 supports the "older" releases.

It looks like the disarm/failsafe bug has now been taken seriously and a PR has been raised with some code changes in it to try and resolve the bug without the need for any major rework of the scheduler code base...

https://github.com/betaflight/betaflight/pull/11354

It looks like the proposal is to try and get this disarm/failsafe bugfix into the Betaflight 4.8.0 RC3 release...

https://github.com/betaflight/betaflight/pull/11355

I was just thinking it may be about time to upgrade to 4.3 so I came here and started reading all this craziness about quads not disarming???? LOLLL ok I'll wait.

(01-Feb-2022, 04:55 AM)the.ronin Wrote: I was just thinking it may be about time to upgrade to 4.3 so I came here and started reading all this craziness about quads not disarming????  LOLLL ok I'll wait.

Yeah probably best to wait. RC3 was released today. Fixes says TBA so dont know what they fixed.
https://github.com/betaflight/betaflight...10.8.0-RC3

Looking at the PR's, it doesn't look like the disarm / failsafe bug made it into the Betaflight 4.3.0 RC3 release. The same code fix to the task scheduler code base appears to have also fixed an issue with the OSD task stalling on some FC's (see HERE) but on the flip side it seems to have caused some other adverse issues with an RXLOSS occurring under some conditions on an FC which has a built in ELRS receiver (see HERE and HERE). So there still seem to be some underlying task scheduler related issues/bugs in the Betaflight 4.3.0 release candidate builds.

There is also some more disagreement amongst the Betaflight developers on another fairly fundamental change that was proposed to be implemented during the release candidate phase that some developers are against, and the other developers state is the wrong approach/solution anyway...

• Move failsafe processing out of the RX task into the scheduler as a real-time task -> https://github.com/betaflight/betaflight/pull/11362
  

So I don't think we are out of the woods yet with some of these potentially critical bugs in the Betaflight 4.3.0 release, and it looks like we are probably in for a few more release candidate builds before we get a final stable release. So personally I am still holding off installing the Betaflight 4.3.0 release candidate builds on any of my quads, and I have already rolled back the only quad I had Betaflight 4.3.0 RC2 installed on to Betaflight 4.2.11.

How long does this typically take? I have a brand new build and just the flashing / tuning remains. It wont be much flying now anyway, still winter here in Sweden but I am wondering if it is worth putting all the effort into tuning the quad.

Following on from post #43, the Betaflight devs have now published the complete list of fixes for the Betaflight 4.3.0 RC3 release and the fixes for the disarm / failsafe bug have actually made it into the RC3 release.

The first post in this thread has now been updated to include the list of fixes for the RC3 release of both the firmware and the configurator.

• Betaflight 4.3 RC3 Firmware & Release Notes: https://github.com/betaflight/betaflight.../4.3.0-RC3
  
• Betaflight Configurator 10.8.0-RC3: https://github.com/betaflight/betaflight...10.8.0-RC3