ERA reports impossible stacktrace

I get a stacktrace like following:

SuntimesUtil.calcData:121
DataField.isContentCenter:141
DataField.initialize:122
DataField.loadHeartRateData:759
WatchFaceView.onSettingsChanged:262

Here's the isContentDataCenter function:

function isContentCenter() {
	return mPos == DataField.POS_TOP_CENTER || mPos == DataField.POS_BOTTOM_CENTER;
}

So how can the stacktrace report that SuntimesUtil.calcData is called from within DataField.isContentCenter? Can the stacktrace be interpreted differently so that t makes sense?

The error type is Error Name: Invalid Value - what does this actually mean?

0 jim_m_58 over 6 years ago

Check the FW version of the device, as if it's running old FW, you can get some odd ERA logs.

0 Michael over 6 years ago in reply to jim_m_58

seems like after an app update, ERA is reset... But got one report already again that looks like following:

Error Name: Invalid Value
Occurrences: 1
First Occurrence: 2020-06-18
Last Occurrence: 2020-06-18
Devices:
    fēnix® 6X Pro / 6X Sapphire / 6X Pro Solar / tactix® Delta Sapphire: 9.00
App Versions: v1.1.3
Languages: deu
Backtrace:
    DataField.loadDatetime:1198
    DataField.loadData:668
    DataField.load:129
    WatchFaceView.onUpdate:137

DataField::loadData 668 looks like following:

	loadSunriseSunset(setup, SuntimesUtil.TodaySunset, false);
	break;

DataField::loadDatetime would be called in 686 (inside another case of the surounding switch)...

The reporting device is running the latest firmware, isn't it? 9.0 seems to be the latest here: www8.garmin.com/.../download_details.jsp

0 jim_m_58 over 6 years ago in reply to Michael

Yes, 9.00 is the latest production FW (should be fine), but there's been a couple of betas since then.

The best place to look is in the forums for a specific device.

https://forums.garmin.com/outdoor-recreation/outdoor-recreation/f/fenix-6-series

0 jim_m_58 over 6 years ago in reply to Michael

at times the line number could be the beginning of a code block where the error is actually within a code block.

0 Michael over 6 years ago in reply to jim_m_58

I'm unsure how you mean this - you mean that if function and line number do not match I should check the function that is called at the reported line number? In my case I should check loadSunriseSunset function. Is this what you mean?

0 jim_m_58 over 6 years ago in reply to Michael

What I'm saying, if if you are doinf something like a for loop, the problem could be in the for loop, and not the for() itself, even if that shows as the line number.

0 tomulli over 6 years ago

I am also getting ERA reports with backtrace which does not make sense (e.g. getting symbol not found on a= b+c linked from function which has nothing to do with it) started approx 3 weeks back. I assume it is Garmin issue and i am ignoring them at the moment.

Based on error and its reported number of occurrences i would expect users to be very vocal if real.

0 li2niu over 4 years ago

Aha, I encounter the same issue