Fw 208 crashed twice during 1 hour flight

[georghg]

I was flying with the strato connected to the IPhone using Flyskyhy today.

after a while strato showed a white page full of text suggesting a reset. After the reset it did not connect to the IPhone properly again (it always needs to be unpaired and newly paired in order to work. The same applies every time after turning on and off).

After working for a while the upper half of the screen had an error only showing lines and dots, while the lower part was still working.

[fhorinek]

Thank you for the report. We will check it out. By any chance can you make a photo of the white page when this happens? Safety first, not try this in flight.

[georghg]

Helle Frantisek,

So far it only happened during flight, but I will try if it also happens at home if I leave it turned on for a while, or maybe in the car.

Kind regards Georg

Von meinem iPhone gesendet

Am 04.01.2022 um 19:55 schrieb František Horínek @.***>:

 Thank you for the report. We will check it out. By any chance can you make a photo of the white page when this happens? Safety first, not try this in flight.

— Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android. You are receiving this because you authored the thread.

[georghg]

Dear Frantisek,

I was able to replicate it in the car. See foto below.

Kind regards Georg

Von meinem iPhone gesendet

Am 04.01.2022 um 19:55 schrieb František Horínek @.***>:

 Thank you for the report. We will check it out. By any chance can you make a photo of the white page when this happens? Safety first, not try this in flight.

— Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android. You are receiving this because you authored the thread.

[georghg]

Dear Frantisek,

I detected another bug. The connection to the Iphone also changes the widgets. In the flight on Sunday I thought I placed the glide ratio twice and forgot the Vario, but in fact the connection to the iphone changed it. In the car I replicated this. See photos below:

Before Bluetooth connection:

After connection to the Iphone:

This happens without restart just by connecting to the iPhone.

Kind regards Georg

Von meinem iPhone gesendet

Am 05.01.2022 um 08:47 schrieb Georg Gottholmseder @.***>:

Dear Frantisek,

I was able to replicate it in the car. See foto below.

Kind regards Georg

Von meinem iPhone gesendet

Am 04.01.2022 um 19:55 schrieb František Horínek @.***>:

 Thank you for the report. We will check it out. By any chance can you make a photo of the white page when this happens? Safety first, not try this in flight.

— Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android. You are receiving this because you authored the thread.

[georghg]

Dear Frantisek,

I tried around a bit and the switch of one widget between vario and glide ratio also happens without Bluetooth. I will observe it a while to find out what triggers the switch. It is not related to the automatic page switching sind I don’t have the glide ratio on that place in any of the pages.

Kind regards Georg

Von meinem iPhone gesendet

Am 05.01.2022 um 11:23 schrieb Georg Gottholmseder @.***>:

Dear Frantisek,

I detected another bug. The connection to the Iphone also changes the widgets. In the flight on Sunday I thought I placed the glide ratio twice and forgot the Vario, but in fact the connection to the iphone changed it. In the car I replicated this. See photos below:

Before Bluetooth connection:

After connection to the Iphone:

This happens without restart just by connecting to the iPhone.

Kind regards Georg

Von meinem iPhone gesendet

Am 05.01.2022 um 08:47 schrieb Georg Gottholmseder @.***>:

Dear Frantisek,

I was able to replicate it in the car. See foto below.

Kind regards Georg

Von meinem iPhone gesendet

Am 04.01.2022 um 19:55 schrieb František Horínek @.***>:

 Thank you for the report. We will check it out. By any chance can you make a photo of the white page when this happens? Safety first, not try this in flight.

— Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android. You are receiving this because you authored the thread.

[fhorinek]

Thank you for your report

I can't see the attached photos, the github will remove them. Please send them to fero@skybean.eu or upload them from the github.

The Glide widget have option to show avg. vario during climb (Since glide ratio in can't be calculated). This behavior can be changed in widget settings in page edit.

[ashghinn]

Today I flew for about 80 minutes and the Strato (fw208) froze four times. I had it connected to my phone via bluetooth, and my phone was running XC Track. I hope the images are clear enough.

[georghg]

[fhorinek]

Thank you for the report, I will check it out

[glaroburger]

have also wuite unstable behaviour. FW208, BLE with XCtrac k

[theosch1]

The frozen display appears again and again and the title of the text that appears is "bus error". The vario beebs at the same rate, although the conditions are changing. The compass does not work properly whereas the heading is correct. In the Fanet settings I can see others, however the distance is always 65 km and the numbers after the @ is unknown to me.

[glaroburger]

Well I run my device on the balcony without any km. I suggest to either to disable any feature step by step and repeat the fly again and again or if you have an idea of the problem, we start with that and try to reproduce.

Von meinem Smartphone gesendet.

-------- Ursprüngliche Nachricht -------- Von: theosch1 @.> Datum: 10.01.22 05:37 (GMT+01:00) An: fhorinek/BB @.> Cc: glaroburger @.>, Comment @.> Betreff: Re: [fhorinek/BB] Fw 208 crashed twice during 1 hour flight (Issue #145)

The frozen display appears again and again and the title of the text that appears is "bus error". The vario beebs at the same rate, although the conditions are changing. The compass does not work properly whereas the heading is correct. In the Fanet settings I can see others, however the distance is always 65 km and the numbers after the @ is unknown to me.

— Reply to this email directly, view it on GitHubhttps://github.com/fhorinek/BB/issues/145#issuecomment-1008539543, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AXHDBPNYQ3EQRHQGKG4GE6DUVJPBTANCNFSM5LEB4GQQ. Triage notifications on the go with GitHub Mobile for iOShttps://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Androidhttps://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub. You are receiving this because you commented.Message ID: @.***>

[glaroburger]

Investigation started and WIP

[fhorinek]

Wow thank you for your effort.

From your investigation, this might be caused by memory leaks caused by telemetry generation (BLE). The memory then corrupt next task memory. I need to dig deeper in the code.

Thank you again

One remark: When the FANET is disabled FLARM is not transmitting. FLARM beacon is part of the FANET module.

[glaroburger]

Or maybe wrong discovery of memory after Interrupt of ble packet rdy?

I did not dig into the code. I'm hardware developer with very basic coding skills...

Thanks for the flarm hint. Fanet on, flarm off ble off still working after 5 hours flight... I will start the ble on, fanet and Flarm off test this night...

Von meinem Smartphone gesendet.

-------- Ursprüngliche Nachricht -------- Von: František Horínek @.> Datum: 10.01.22 19:15 (GMT+01:00) An: fhorinek/BB @.> Cc: glaroburger @.>, Comment @.> Betreff: Re: [fhorinek/BB] Fw 208 crashed twice during 1 hour flight (Issue #145)

Wow thank you for your effort.

From your investigation, this might be caused by memory leaks caused by telemetry generation (BLE). The memory then corrupt next task memory. I need to dig deeper in the code.

Thank you again

One remark: When the FANET is disabled FLARM is not transmitting. FLARM beacon is part of the FANET module.

— Reply to this email directly, view it on GitHubhttps://github.com/fhorinek/BB/issues/145#issuecomment-1009196073, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AXHDBPP5S2C4FW4XB4EE463UVMOYPANCNFSM5LEB4GQQ. Triage notifications on the go with GitHub Mobile for iOShttps://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Androidhttps://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub. You are receiving this because you commented.Message ID: @.***>

[fhorinek]

The real transfer via BLE is done in second (ESP32) processor so the memory overwrite/leak must occur during the request for telemetry transfer inside of timer task.

The error says that cpu is in invalid state. Normally this can only happen during (incorrect) programming in assembly. My best guess is that the string buffers overflow/underflow and corrupt the stack for next scheduled task. The cpu registers are reloaded with bad data and that triggers the fault.

Can try to use classic BT for next test? This way we can isolate the bug inside the telemetry formatting routines.

[glaroburger]

Fanet on only works now for 5h37. I will start the classic BT test now.

As a minor request. Could you consider to group the Flarm function as a supgroup of Fane, so that it’s clearly visible that Flarm is not working when Fanet is off?

If you want to be less trackable, maybe someone thinks of using Flarm only without transmitting to Fanet

Von: František Horínek @.> Gesendet: Montag, 10. Januar 2022 21:38 An: fhorinek/BB @.> Cc: glaroburger @.>; Comment @.> Betreff: Re: [fhorinek/BB] Fw 208 crashed twice during 1 hour flight (Issue #145)

The real transfer via BLE is done in second (ESP32) processor so the memory overwrite/leak must occur during the request for telemetry transfer inside of timer task.

The error says that cpu is in invalid state. Normally this can only happen during (incorrect) programming in assembly. My best guess is that the string buffers overflow/underflow and corrupt the stack for next scheduled task. The cpu registers are reloaded with bad data and that triggers the fault.

Can try to use classic BT for next test? This way we can isolate the bug inside the telemetry formatting routines.

— Reply to this email directly, view it on GitHubhttps://github.com/fhorinek/BB/issues/145#issuecomment-1009322431, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AXHDBPM6N5BHDCFCEUCIIU3UVM7UHANCNFSM5LEB4GQQ. Triage notifications on the go with GitHub Mobile for iOShttps://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Androidhttps://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub. You are receiving this because you commented.Message ID: @.**@.>>

[glaroburger]

Flarm/Fanet OFF; Blue-tooth SPP tele on, A2DP audio off an GNSS forwarding off works without Bus error over 9hours. The Error is associated with the GNSS forwarding.

With GNSS forwarding ON, I had 2 crashes within a short time.

[fhorinek]

This is excellent catch. Unfortunately I am abusing the gnss forward function, but no luck with the crash. Do you have igc or log files from the crash? if so send them to me to fero@skybean.eu

[georghg]

I am at work right now and don‘t have the strato here with me. But I should still have my igc file on the device and could post it in the evening.

[Altgenug1]

Hy, I had a one hour flight today. Immediately after starting, the Strato froze with an error message. I was able to turn it back on with the reset button. Attached you will find the debug.log file. My Strato was also connected to the iPhone and FlySkyHy via Bluetooth. I hope the information helps a little. debug.log