* Now talking on #necessitas
<tomaslov> bog_dan_ro: ok, now in linux - ready to test
<bog_dan_ro> tomaslov: go to qtbase
<tomaslov> bog_dan_ro: yep
<bog_dan_ro> copy https://paste.kde.org/pbslwlwa9 to a file
<bog_dan_ro> $ patch -p1 < file.diff
<bog_dan_ro> make
<bog_dan_ro> then rerun the application and collect the log
<bog_dan_ro> rebuild (with the new qt libs) & rerun the application :)
* tomaslov has quit (Ping timeout: 265 seconds)
* Disconnected (Remote host closed socket)
* tomaslov sets mode +Z on tomaslov
* tomaslov sets mode +i on tomaslov
* Now talking on #necessitas
* lqi has quit (Ping timeout: 250 seconds)
* ylopes has quit (Ping timeout: 260 seconds)
<tomaslov> bog_dan_ro: make done, just struggling to get the device connected with adb now.. 
* anshaw has quit (Ping timeout: 276 seconds)
<bog_dan_ro> tomaslov: try $ adb kill-server; sudo adb start-server
<bog_dan_ro> most probably you don't updated your udev permissions
* anshaw (~quassel@185.55.107.82) has joined
<tomaslov> bog_dan_ro: that helped
* ylopes (~ylopes@185.55.107.82) has joined
<tomaslov> bog_dan_ro: I can see 08-23 14:30:09.864: E/JavaBinder(2933): !!! FAILED BINDER TRANSACTION !!!
<tomaslov> bog_dan_ro:  will upload the whole adb logcat now..
<tomaslov> bog_dan_ro: https://paste.kde.org/p8zyvxsyx
<bog_dan_ro> tomaslov: I think I spot somehting, I need you to change one more thing to qtbase
<bog_dan_ro> tomaslov: open src/plugins/platforms/android/androidjnimain.cpp
<bog_dan_ro> and in "
<bog_dan_ro> static void setSurface(JNIEnv *env, jobject /*thiz*/, jint id, jobject jSurface, jint w, jint h)
<bog_dan_ro> {
<bog_dan_ro> "
<bog_dan_ro> add
<bog_dan_ro> qWarning()<<"setSurface" << id << jSurface << w << h;
<bog_dan_ro> as first line
<tomaslov> bog_dan_ro: sure, I'll change this
<bog_dan_ro> tomaslov: build qtbase & rebuild & rerun the application & paste the log again
<bog_dan_ro> qWarning()<<"setSurface" << id << (void*)jSurface << w << h;
<bog_dan_ro> is better
<bog_dan_ro> tomaslov: ^
<tomaslov> bog_dan_ro: ok.
<tomaslov> bog_dan_ro:  https://paste.kde.org/ph6y8fyru
<tomaslov> bog_dan_ro: 08-23 14:46:19.711: W/libhelloworld3.so(9491): androidjnimain.cpp:588 (void setSurface(JNIEnv*, jobject, jint, jobject, jint, jint)): setSurface 1 0x0 0 0
* paulot (~paul@185.55.107.82) has joined
<bog_dan_ro> tomaslov: that's ok
<bog_dan_ro> tomaslov: try this, start the application, press home button and then start it again from apps menu
<bog_dan_ro> tomaslov: if it doesn't crash do it again a few times then give me the log
<tomaslov> bog_dan_ro: ok, will test
<tomaslov> bog_dan_ro: it did crash the second time: https://paste.kde.org/pyd7p5o4c
* lqi (~lqi@qt/lqi) has joined
<bog_dan_ro> tomaslov: what android version the device has ?
<bog_dan_ro> found it
<bog_dan_ro> 5.1.1
<tomaslov> yeps
* g10 has quit (Ping timeout: 260 seconds)
<tomaslov> Qt5.6+ and 5.1.1 are not good friends?
<bog_dan_ro> tomaslov: it seems we changed something that 5.1.1 doesn't like ... but I have no idea what...
<bog_dan_ro> tomaslov: sadly in this moment I have no idea what's wrong ... we'll need a way to reproduce it on an emulator
<tomaslov> bog_dan_ro: if it's device-specific, or related to the GL hardware, that could be quite tricky
<bog_dan_ro> tomaslov: yeah ...
<bog_dan_ro> eskil: folks do you have such device ?
<eskil> Just catching up
<eskil> We don't have that particular one, but I can check the specs
<bog_dan_ro> eskil: any device that we can reproduce the issue will be fine
<bog_dan_ro> eskil: If you want to test something else tomaslov has the device ...
<tomaslov> eskil: I still have the device for 15 more minutes.. :)
<bog_dan_ro> I thought the java part is reinitilized and that was the reason why it crashes but is not the case
<eskil> We have a SM-T110 which also has marvell + vivante, but older chips and Android 4
* g10 (~kvirc@92.185.225.241) has joined
<tomaslov> eskil: we've now experienced the bug with just: 1. start a helloworld Qt Quick Controls app, 2. press home, 3. restart the app
* ZaggeO has quit (Remote host closed the connection)
* lqi_ (~lqi@qt/lqi) has joined
* lqi has quit (Ping timeout: 250 seconds)
* anshaw has quit (Ping timeout: 265 seconds)
<tomaslov> eskil, bog_dan_ro: the customer has now picked up his  device here, so I'm without hardware to test this on as well. Please give me a hint if it could be of any help if I ask some other user to lend me one again. 
<eskil> tomaslov: Is it reproducible every time with the recipe above?
<tomaslov> eskil, not every time - but about 1 in 2
<tomaslov> which is consistent with the results from the recipe described in the bugreport