tgkill+12

Are this all from same version 2018.3350?

Yes, all this crashes are from version 2018.3350

Regards

Hello!

This look much better! Here are symbolicated crashes. What happens here is following:

  1. Update Display::Update(), which dispatches “enterFrame” event.

  2. Then it executes the event and error during invocation or executing of the event.

Enter Frame can have many things in it, it’s a workhorse of things like transitions or timers, so error may happen like inside a timer or transition, trying to access deleted widget or display object or something.

So this error is caused by some Lua code, and should be reproducible. Also, it is weird, because usually Lua errors should not crash the app. Do you use some custom error handling?

Great we are closer… here are 3 questions:

1-We recently added this block of code trying to catch the error, but unfortunely we have no luck. We have not received any crash report from the app. Is there any chance that this error handler is hinding the cause of the error?

2-Is there any way to get the lua filename to find in wich screen is crashing?

3-Did you recommend that i try to remove every enterframe handler replacing it by transitions or this error can be caused by a timer to?

Thanks! 

this is the error handler that we added:

local function handleUnexpectedError(event)

analyticsManager.logEvent( “Error”, 

      {

         errorMessage=event.errorMessage, 

         stackTrace=event.stackTrace

      } 

   )

end

Runtime:addEventListener(“unhandledError”, handleUnexpectedError)

  1. There’s misunderstanding: I told that transitions and times utilize enterFrame within them. So the cause may be not your direct enterFrame handler, but some of timers.

  2. Did you add that code after the crashes appear or before? I would suggest testing it first somehow. Errors in error handler is a nasty thing. Also, if your handler invoker network request it may cause problems.

  3. Not on google console. It is hard to tell what is going on inside your Lua code.If you would have a repro for crash, you would see it in logcat output

Thanks for your help Vlads, we found an issue with a timer handler in the new mini game. I hope that this is the issue that was looking for. We already released a new version, so lets wait and see.

It would be great if someday we can get the stacktrace in play console just like we have in simulator, logcat, etc.

Best Regards

Can you provide us with the version number of Corona you’re using (please don’t say “The Latest”, but actually get the number from the “About Corona” menu link or from the welcome window)?

Are you building on macOS or Windows?

Hello Rob, im using the version 2018.3443 and building on Windows 10 laptop.

Regars

It is hard to tell what is going on from the traceback, it seems that image decoding is faulting for some reason. Crash happens when native corona library tries to locate and decode image using java call, and it fails to find a java call for mysterious reasons. It should not happen.

  1. What images are you using? Is it possible you’re using downloaded images or something like that?

  2. Are there particular devices/manufacturers/android versions associated with the crash?

  3. You say “a lot of crashes” but how many is that, in percents, not absolute numbers.

Hello Vlads, here are my responses:

1-I’m using png images, the same images in all devices. All images was created by our design team and they are included in the apk file, we do not download any image.

2-There is no particular devices. See attachments.

3-In percentage it represents about 329% more crashes than previous version. See the attachments.

Is there any way to catch this exceptions and log in somewere the stacktrace to indentify the problem?

Thanks for your help.

Best regards

This is the errors trend:

errors.jpg

This is the device list of one instance of the crash:

devices.jpg

Hello again, any update? Is there anything that I can do to resolve the issue?

We are loosing positions on play store because this errors.

Thanks

Unfortunately it’s not much to go with: I decoded stack-trace and it does not point to any reasonable error. All I see is that decoder attempts to decode “.png” file. Actually, it crashes when it tries to locate it inside APK achieve. Thing is, this operation is done with every sprite you load in every android game, and directly related code wasn’t changed forever (like last 3-4 years or so). I would love to help, but I am at loss - I don’t see how this can be happening and I don’t have a way to reproduce the bug. I really want to help, but I don’t know how…

What changed on November 29, so you started to get crashes? Did you move to new version of Corona? Which was previous version? Is it possible to use that one?

On november 25 we released a new version of the app. It is built under the new version of Corona (2018.3443) and also including a new mini game in the app.

So i’m not sure if the problem is caused by the corona sdk upgrade or some image in the new game.

Did you recomend that I downgrade to the previous version of corona SDK to see the if the error persists?

Is there any way to catch the exception to identify the lua file or the image name to see this particular element?

Regads

You could put something in build.settings to prevent us from stripping debugging information. See:

http://docs.coronalabs.com/guide/distribution/advancedSettings/index.html#build-control

Generally, you don’t want to change multiple things when tracking down a problem like this. But dropping back to the previous version of Corona that you built with would be a logical step.

Rob

Rob, it seems like crash happens when Corona loads textures, which happens between frames, not in direct response to Lua code. Crash there should not happen, unless something goes really odd (which unfortunately the case here).

What version of Corona your previous release uses? If it’s older than 2018.3326 I would recommend to roll back Corona version to see if it would help. Note, that Google Play console has latency of few days when crashes are fixed.

Hey guys, we downgraded to the previous version of corona sdk (2018.3350) and released a new version with the flag neverStripDebugInfo = true

The problem still there, here is some of the crashes reported in play developer console.

Any clue?

Thanks!

pid: 0, tid: 0 >>> com.blackbean.tinypuzzle <<< backtrace:

 

#00 pc 000000000004b3ac /system/lib/libc.so (tgkill+12)

 

#01 pc 000000000001a473 /system/lib/libc.so (abort+54)

 

#02 pc 000000000033a8bb /system/lib/libart.so (_ZN3art7Runtime5AbortEPKc+370)

 

#03 pc 000000000033aeb3 /system/lib/libart.so (_ZN3art7Runtime7AborterEPKc+10)

 

#04 pc 00000000003f3419 /system/lib/libart.so (_ZN7android4base10LogMessageD1Ev+456)

 

#05 pc 000000000035d0a5 /system/lib/libart.so (_ZNK3art6Thread24AssertNoPendingExceptionEv+844)

 

#06 pc 00000000000eb961 /system/lib/libart.so (_ZN3art11ClassLinker9FindClassEPNS_6ThreadEPKcNS_6HandleINS_6mirror11ClassLoaderEEE+28)

 

#07 pc 000000000025cc97 /system/lib/libart.so (_ZN3art3JNI9FindClassEP7_JNIEnvPKc+1066)

 

#08 pc 0000000000031774 /data/app/com.blackbean.tinypuzzle–oDfUmsfX2-zwsN9zGpRwg==/lib/arm/libcorona.so

 

#09 pc 00000000001068bc /data/app/com.blackbean.tinypuzzle–oDfUmsfX2-zwsN9zGpRwg==/lib/arm/libcorona.so

 

#10 pc 000000000000cc1c /data/app/com.blackbean.tinypuzzle–oDfUmsfX2-zwsN9zGpRwg==/lib/arm/liblua.so

 

#11 pc 000000000000d01c /data/app/com.blackbean.tinypuzzle–oDfUmsfX2-zwsN9zGpRwg==/lib/arm/liblua.so

 

#12 pc 000000000000c084 /data/app/com.blackbean.tinypuzzle–oDfUmsfX2-zwsN9zGpRwg==/lib/arm/liblua.so

 

#13 pc 000000000000c0e0 /data/app/com.blackbean.tinypuzzle–oDfUmsfX2-zwsN9zGpRwg==/lib/arm/liblua.so

 

#14 pc 000000000000c188 /data/app/com.blackbean.tinypuzzle–oDfUmsfX2-zwsN9zGpRwg==/lib/arm/liblua.so

 

#15 pc 000000000000ca6c /data/app/com.blackbean.tinypuzzle–oDfUmsfX2-zwsN9zGpRwg==/lib/arm/liblua.so

 

#16 pc 000000000001ce30 /data/app/com.blackbean.tinypuzzle–oDfUmsfX2-zwsN9zGpRwg==/lib/arm/liblua.so

 

#17 pc 000000000000d068 /data/app/com.blackbean.tinypuzzle–oDfUmsfX2-zwsN9zGpRwg==/lib/arm/liblua.so

 

#18 pc 000000000000c374 /data/app/com.blackbean.tinypuzzle–oDfUmsfX2-zwsN9zGpRwg==/lib/arm/liblua.so

 

#19 pc 000000000000d1e0 /data/app/com.blackbean.tinypuzzle–oDfUmsfX2-zwsN9zGpRwg==/lib/arm/liblua.so

 

#20 pc 00000000000055b8 /data/app/com.blackbean.tinypuzzle–oDfUmsfX2-zwsN9zGpRwg==/lib/arm/liblua.so (lua_pcall+88)

 

#21 pc 0000000000107070 /data/app/com.blackbean.tinypuzzle–oDfUmsfX2-zwsN9zGpRwg==/lib/arm/libcorona.so

 

#22 pc 000000000009ade4 /data/app/com.blackbean.tinypuzzle–oDfUmsfX2-zwsN9zGpRwg==/lib/arm/libcorona.so

 

#23 pc 0000000000141244 /data/app/com.blackbean.tinypuzzle–oDfUmsfX2-zwsN9zGpRwg==/lib/arm/libcorona.so

 

#24 pc 000000000008d4f8 /data/app/com.blackbean.tinypuzzle–oDfUmsfX2-zwsN9zGpRwg==/lib/arm/libcorona.so

 

#25 pc 000000000008555f /data/app/com.blackbean.tinypuzzle–oDfUmsfX2-zwsN9zGpRwg==/oat/arm/base.odex

Another instance of the same error:

*** *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** pid: 0, tid: 0 >>> com.blackbean.tinypuzzle <<< backtrace:

 

#00 pc 0000000000049c44 /system/lib/libc.so (tgkill+12)

 

#01 pc 00000000000473e3 /system/lib/libc.so (pthread_kill+34)

 

#02 pc 000000000001d535 /system/lib/libc.so (raise+10)

 

#03 pc 0000000000019081 /system/lib/libc.so (__libc_android_abort+34)

 

#04 pc 00000000000170e4 /system/lib/libc.so (abort+4)

 

#05 pc 0000000000319071 /system/lib/libart.so (_ZN3art7Runtime5AbortEv+252)

 

#06 pc 00000000000b4979 /system/lib/libart.so (_ZN3art10LogMessageD2Ev+864)

 

#07 pc 0000000000332623 /system/lib/libart.so (_ZNK3art6Thread24AssertNoPendingExceptionEv+602)

 

#08 pc 00000000000e4679 /system/lib/libart.so (_ZN3art11ClassLinker9FindClassEPNS_6ThreadEPKcNS_6HandleINS_6mirror11ClassLoaderEEE+28)

 

#09 pc 000000000026017f /system/lib/libart.so (_ZN3art3JNI9FindClassEP7_JNIEnvPKc+1878)

 

#10 pc 0000000000031774 /data/app/com.blackbean.tinypuzzle-1/lib/arm/libcorona.so

 

#11 pc 00000000001068bc /data/app/com.blackbean.tinypuzzle-1/lib/arm/libcorona.so

 

#12 pc 000000000000cc1c /data/app/com.blackbean.tinypuzzle-1/lib/arm/liblua.so

 

#13 pc 000000000000d01c /data/app/com.blackbean.tinypuzzle-1/lib/arm/liblua.so

 

#14 pc 000000000000c084 /data/app/com.blackbean.tinypuzzle-1/lib/arm/liblua.so

 

#15 pc 000000000000c0e0 /data/app/com.blackbean.tinypuzzle-1/lib/arm/liblua.so

 

#16 pc 000000000000c188 /data/app/com.blackbean.tinypuzzle-1/lib/arm/liblua.so

 

#17 pc 000000000000ca6c /data/app/com.blackbean.tinypuzzle-1/lib/arm/liblua.so

 

#18 pc 000000000001ce30 /data/app/com.blackbean.tinypuzzle-1/lib/arm/liblua.so

 

#19 pc 000000000000d068 /data/app/com.blackbean.tinypuzzle-1/lib/arm/liblua.so

 

#20 pc 000000000000c374 /data/app/com.blackbean.tinypuzzle-1/lib/arm/liblua.so

 

#21 pc 000000000000d1e0 /data/app/com.blackbean.tinypuzzle-1/lib/arm/liblua.so

 

#22 pc 00000000000055b8 /data/app/com.blackbean.tinypuzzle-1/lib/arm/liblua.so (lua_pcall+88)

 

#23 pc 0000000000107070 /data/app/com.blackbean.tinypuzzle-1/lib/arm/libcorona.so

 

#24 pc 000000000009ade4 /data/app/com.blackbean.tinypuzzle-1/lib/arm/libcorona.so

 

#25 pc 0000000000141244 /data/app/com.blackbean.tinypuzzle-1/lib/arm/libcorona.so

 

#26 pc 000000000008d4f8 /data/app/com.blackbean.tinypuzzle-1/lib/arm/libcorona.so

 

#27 pc 00000000006581dd /data/app/com.blackbean.tinypuzzle-1/oat/arm/base.odex

Hello again, have you checked my latest post? is there anything that point to the error?

Thanks

Yes, it points to the identical nondescript error as before. But since you reverted version of Corona it is plausible to assume crash is caused by some new introduced feature or plugins. What plugins are you using, and when was last non-crashy release of your game?

Hello Vlads, we do not include any new plugin. Just a new mini-game. In that case how can i know wich image/file/whatever is causing the crash? is my first app in corona and it is a new world for me.

Thanks.

May be we updated some plugin you’re using. Please, can you use https://pastebin.com/ to share your build.settings?

Also, this is not a nice situation. Clearly, Corona should not crash in any circumstances. But I struggle to understand what is going on. Try to test changed parts of code etc.