*** iTune has joined #sailfishos-porters | 00:01 | |
*** mispp has quit IRC | 00:25 | |
*** Core2idiot has joined #sailfishos-porters | 00:34 | |
*** olafh has quit IRC | 00:35 | |
*** Core2idiot has quit IRC | 00:40 | |
*** iTune has quit IRC | 01:11 | |
*** Core2idiot has joined #sailfishos-porters | 01:11 | |
*** Core2idiot has quit IRC | 01:18 | |
*** DaZ has quit IRC | 01:21 | |
*** vfrgbgf has joined #sailfishos-porters | 01:22 | |
*** Zotan has quit IRC | 01:23 | |
vfrgbgf | Why do I use sailfishOS mic created not start? | 01:24 |
---|---|---|
*** marxistvegan has quit IRC | 01:30 | |
*** DaZ has joined #sailfishos-porters | 01:35 | |
*** marxistvegan has joined #sailfishos-porters | 01:35 | |
vfrgbgf | what's time? | 01:43 |
vfrgbgf | Now 1:43AM, no one looks. | 01:44 |
vfrgbgf | I want someone to solve my problem. | 01:46 |
vfrgbgf | locusf,Are you there? | 01:48 |
vfrgbgf | lbt, sledges,Are you there? | 01:53 |
vfrgbgf | Some successful porters, Are you there? | 01:59 |
vfrgbgf | alterego,Are you there? | 02:01 |
*** marxistvegan has quit IRC | 02:04 | |
*** vfrgbgf has quit IRC | 02:07 | |
*** iTune has joined #sailfishos-porters | 02:10 | |
*** PhompAng has joined #sailfishos-porters | 02:20 | |
*** iTune has quit IRC | 02:21 | |
*** lpotter has quit IRC | 02:23 | |
*** lpotter has joined #sailfishos-porters | 02:23 | |
*** PhompAng has quit IRC | 02:44 | |
*** zhxt_work has quit IRC | 02:44 | |
*** PhompAng has joined #sailfishos-porters | 02:47 | |
*** zhxt_work has joined #sailfishos-porters | 02:57 | |
*** PhompAng has quit IRC | 03:00 | |
*** marxistvegan has joined #sailfishos-porters | 03:05 | |
*** marxistvegan has quit IRC | 03:14 | |
*** marxistvegan has joined #sailfishos-porters | 03:14 | |
*** PhompAng has joined #sailfishos-porters | 03:21 | |
*** Umeaboy has joined #sailfishos-porters | 03:52 | |
Umeaboy | sledges: You awake? :) | 04:05 |
*** marxistvegan has quit IRC | 04:22 | |
*** iGhost has joined #sailfishos-porters | 04:30 | |
*** furikku has joined #sailfishos-porters | 04:30 | |
*** PhompAng has quit IRC | 04:34 | |
*** iGhost has quit IRC | 05:19 | |
*** filippz has joined #sailfishos-porters | 05:36 | |
*** filippz has quit IRC | 05:44 | |
*** olafh has joined #sailfishos-porters | 05:47 | |
_dinsdale | Hello, I am working through the hadk on section 7.1.1 Building the droid-hal-device packages... | 05:54 |
_dinsdale | I am running into the issue outlined in my pastebin at http://pastebin.com/0aM059UF | 05:55 |
_dinsdale | Stskeeps pointed out that I need to modify the file cyanogenmod_bacon_defconfig as specified in ../android/droid/device/oneplus/bacon/BoardConfig.mk | 05:56 |
_dinsdale | I have updated that file, but I continue to get the same CONFIG warnings as before. Is there possibly a second place where this information would be specified? | 05:57 |
_dinsdale | I have modified the file ../android/droid/kernel/oneplus/msm8974/arch/arm/configs/cyanogenmod_bacon_defconfig | 05:57 |
_dinsdale | I am trying to build for a OnePlus One | 05:59 |
_dinsdale | (bacon) | 05:59 |
*** filippz has joined #sailfishos-porters | 05:59 | |
Umeaboy | _dinsdale: Seems like you have the same problem as I did a while ago. | 05:59 |
_dinsdale | do tell! | 06:00 |
Umeaboy | I'd create the pattern files and make a pull request. | 06:02 |
Umeaboy | https://github.com/mer-hybris/droid-hal-device/tree/hybris-10.1/patterns | 06:02 |
Umeaboy | I don't see any for OnePlus One there. | 06:03 |
Umeaboy | Make a fork of https://github.com/mer-hybris/droid-hal-device/tree/hybris-10.1/patterns/templates | 06:03 |
Umeaboy | and change thoose files and then make a pull request. | 06:03 |
Umeaboy | I don't remember how many lines you have to go back, but that would solve it. | 06:08 |
locusf | _dinsdale: did you rebuild your kernel after the config changes? | 06:10 |
Umeaboy | Did you also remember to do rm .repo/local_manifests/roomservice.xml | 06:12 |
Umeaboy | after the breakfast bacon command? | 06:12 |
_dinsdale | no, I did not go back that far. | 06:12 |
_dinsdale | So is breakfast building my kernel? | 06:13 |
_dinsdale | (sorry all this is new to me) | 06:14 |
_dinsdale | ah, yes dinsdale, if you READ the text that accompanies the commands in the HADK... lolz | 06:14 |
_dinsdale | okay, I will try that | 06:15 |
_dinsdale | Umeaboy: do I need to fork the tree directory? I just created a new .spec file hoping that would be enough | 06:16 |
Umeaboy | _dinsdale: That would help for the future and nobody has ported to that device yet......at least not that I know anyway ;) | 06:17 |
_dinsdale | Mr. vgrades apparently has done some work on it. Okay, I'll look into that. Thanks for the advice | 06:18 |
*** iTune has joined #sailfishos-porters | 06:18 | |
Umeaboy | You're welcome. ;) | 06:19 |
*** VDVsx has quit IRC | 06:22 | |
*** VDVsx has joined #sailfishos-porters | 06:35 | |
_dinsdale | What is the best forking strategy for this project? should I just fork /github.com/mer-hybris/droid-hal-device or should I for the whole damn thing? | 06:56 |
Umeaboy | _dinsdale: Please do as I told you. | 06:58 |
Umeaboy | Only fork the template and change thoose files and make a pull request. | 06:58 |
_dinsdale | okay, thanks. :) | 07:00 |
Umeaboy | I'm kinda stupid so I just did that for my model and then I waited until the pull request was merged into master and then I started over completely. | 07:00 |
Umeaboy | ......In the HADK. | 07:00 |
_dinsdale | oh, okay. Chapter? | 07:01 |
Umeaboy | I have even made a slim version of it myself that I use everytime. | 07:01 |
_dinsdale | I've read the thing over and over again but it was back a few weeks now | 07:01 |
Umeaboy | 4.1 | 07:01 |
_dinsdale | because it's all so knew I get turned upside down often. :) | 07:01 |
Umeaboy | 4.1.2 if I'm not misstaken. | 07:01 |
Umeaboy | That's when we all learn something good. | 07:02 |
Umeaboy | To never do that to often. | 07:02 |
_dinsdale | I've wound up getting this far almost despite my best efforts. THe handbook has been excellent. | 07:02 |
*** iTune has quit IRC | 07:03 | |
_dinsdale | ah, Chapter 14 | 07:10 |
_dinsdale | okay, great. Thats it for me tonight. Tootles! :D | 07:11 |
*** _dinsdale has quit IRC | 07:15 | |
*** eleroux has joined #sailfishos-porters | 07:26 | |
*** Snepsts has quit IRC | 07:29 | |
*** gogeta_ has joined #sailfishos-porters | 07:37 | |
*** carepack_ has quit IRC | 07:45 | |
*** Sequenced has joined #sailfishos-porters | 07:46 | |
*** arcean has joined #sailfishos-porters | 07:48 | |
*** gogeta_ has quit IRC | 07:49 | |
*** iTune has joined #sailfishos-porters | 08:03 | |
*** PhompAng has joined #sailfishos-porters | 08:04 | |
*** gogeta_ has joined #sailfishos-porters | 08:06 | |
*** iTune has quit IRC | 08:08 | |
*** Tassadar has joined #sailfishos-porters | 08:12 | |
*** filippz has quit IRC | 08:16 | |
*** filippz has joined #sailfishos-porters | 08:16 | |
*** gogeta_ has quit IRC | 08:36 | |
*** Nokius has quit IRC | 09:12 | |
*** Nokius has joined #sailfishos-porters | 09:12 | |
*** alin has joined #sailfishos-porters | 09:21 | |
*** vrutkovs has joined #sailfishos-porters | 09:31 | |
*** Zotan has joined #sailfishos-porters | 09:37 | |
*** Zotan has quit IRC | 09:58 | |
*** morphis has joined #sailfishos-porters | 10:02 | |
*** vrutkovs has quit IRC | 10:05 | |
*** carepack has joined #sailfishos-porters | 10:13 | |
*** Nokius1 has joined #sailfishos-porters | 10:26 | |
*** Nokius has quit IRC | 10:28 | |
*** blackjack4it has joined #sailfishos-porters | 10:28 | |
*** blackjack4it has quit IRC | 10:29 | |
*** vrutkovs has joined #sailfishos-porters | 10:29 | |
*** spiiroin has joined #sailfishos-porters | 10:29 | |
*** alin has quit IRC | 10:33 | |
*** alin has joined #sailfishos-porters | 10:56 | |
*** iTune has joined #sailfishos-porters | 11:07 | |
*** PhompAng has quit IRC | 11:09 | |
*** iTune has quit IRC | 11:15 | |
*** alin has quit IRC | 11:24 | |
*** cxl000 has joined #sailfishos-porters | 11:29 | |
*** guanche has joined #sailfishos-porters | 11:32 | |
*** filippz has quit IRC | 11:39 | |
*** filippz has joined #sailfishos-porters | 11:39 | |
*** vrutkovs has quit IRC | 11:59 | |
*** vrutkovs has joined #sailfishos-porters | 12:00 | |
*** vakkov has quit IRC | 12:19 | |
MSameer | energycsdx: no difference. both don't have camera working :p | 12:19 |
*** alin has joined #sailfishos-porters | 12:21 | |
energycsdx | MSameer: and between n4 and sbj? | 12:33 |
MSameer | energycsdx: what are you looking for? | 12:35 |
MSameer | I may help more if I know | 12:35 |
energycsdx | MSameer: just investigating to what i have to do to have camera working | 12:36 |
MSameer | which phone do you have? | 12:36 |
*** vakkov has joined #sailfishos-porters | 12:48 | |
*** iTune has joined #sailfishos-porters | 12:59 | |
*** vakkov has quit IRC | 13:00 | |
*** guanche has quit IRC | 13:01 | |
*** vakkov has joined #sailfishos-porters | 13:05 | |
energycsdx | MSameer: Xperia L, MSM8230 | 13:11 |
*** cl has joined #sailfishos-porters | 13:16 | |
MSameer | I have no idea about that | 13:16 |
MSameer | basically either you try the existing stuff and see if it works for you | 13:17 |
MSameer | and if not, write the missing code (depends on what's missing) | 13:17 |
MSameer | or wait for gst 1.0 (a few months) and you should hopefully get camera working with minimal effort | 13:17 |
MSameer | just tell me which approach you fancy and I will help you more as much as I can | 13:18 |
*** PhompAng has joined #sailfishos-porters | 13:20 | |
*** iTune has quit IRC | 13:24 | |
*** energycsdx has quit IRC | 13:25 | |
*** energycsdx_ has joined #sailfishos-porters | 13:25 | |
energycsdx_ | MSameer: what can be missing? i thought camera API is same on different devices | 13:26 |
MSameer | no it's not :) | 13:26 |
MSameer | the high level API is but not the low level ones | 13:27 |
*** alin has quit IRC | 13:30 | |
energycsdx_ | MSameer: how to check which lowlevel API i have? | 13:31 |
MSameer | do you have the gstreamer plugins installed? | 13:32 |
MSameer | gstreamer0.10-droidcamsrc | 13:32 |
energycsdx_ | yes | 13:32 |
*** Tassadar_ has joined #sailfishos-porters | 13:33 | |
energycsdx_ | MSameer: camera app starts, but preview is black | 13:34 |
MSameer | try: | 13:34 |
MSameer | GST_DEBUG='droidcam:5' gst-launch-0.10 -e droidcamsrc | 13:35 |
*** Tassadar has quit IRC | 13:37 | |
MSameer | this should warn us if the camera hal is not supported and give the version number | 13:38 |
*** Sequenced has quit IRC | 13:46 | |
*** Umeaboy has quit IRC | 13:49 | |
*** vakkov has quit IRC | 13:54 | |
energycsdx_ | MSameer: i don`t see any version number | 13:55 |
energycsdx_ | it complains: "streaming task paused, reason not-linked(1)" | 13:55 |
energycsdx_ | -1 | 13:56 |
*** energycsdx_ has quit IRC | 14:08 | |
*** eleroux has quit IRC | 14:08 | |
*** energycsdx has joined #sailfishos-porters | 14:09 | |
*** marxistvegan has joined #sailfishos-porters | 14:10 | |
carepack | hi guys | 14:10 |
energycsdx | hi | 14:10 |
carepack | my problem from yesterday is still present. How to setup the repo for gstreamer-coloconv? | 14:12 |
carepack | the mic create command fails -> gstreamer is missing | 14:13 |
carepack | energycsdx: you gave me some hint yesterday. Treid to comment gstreamer out did not help | 14:13 |
energycsdx | carepack: after you commented out, you should apply this change to your repo | 14:17 |
*** filippz has quit IRC | 14:19 | |
carepack | this means: step 7.1.2 Create a local RPM repository - again | 14:19 |
*** vakkov has joined #sailfishos-porters | 14:25 | |
energycsdx | no | 14:25 |
energycsdx | carepack: 8.3 | 14:25 |
carepack | ah. thx. But did this already and the dependency is still there. but it's another day. So I'm trying again | 14:26 |
energycsdx | carepack: i suggest to remove .repo dir | 14:27 |
MSameer | energycsdx: full output please | 14:28 |
MSameer | sledges: can you help with this camera business ? :) ^ | 14:28 |
energycsdx | MSameer: http://pastebin.com/kdMnjg0q | 14:34 |
carepack | also when I comment out everything related to gstream (gstreamen in the name, pulseaudio) he expect them. because it's a dependency of jolla-hw-adaptation-mak. marking pattern, jolla-configuratin-mako | 14:37 |
*** vakkov has quit IRC | 14:52 | |
*** vakkov has joined #sailfishos-porters | 14:57 | |
MSameer | energycsdx: try gst-launch-0.10 -e droidcamsrc ! droideglsink | 15:01 |
MSameer | energycsdx: if that works then your camera should work | 15:02 |
MSameer | what would be missing is some configuration for jolla-camera which sledges can help you with ;-) | 15:02 |
energycsdx | MSameer: it segfaulted | 15:03 |
*** PhompAng has quit IRC | 15:06 | |
*** PhompAng has joined #sailfishos-porters | 15:07 | |
*** alin has joined #sailfishos-porters | 15:10 | |
MSameer | energycsdx: have you built those elements against your android hal? | 15:15 |
energycsdx | yes i built it by myself | 15:16 |
*** viper0984 has joined #sailfishos-porters | 15:16 | |
energycsdx | should i run gst-launch-0.10 -e droidcamsrc ! droideglsink from fingerterm or ssh is ok? | 15:17 |
*** beidl_ has joined #sailfishos-porters | 15:20 | |
*** beidl has quit IRC | 15:23 | |
MSameer | energycsdx: i always use ssh | 15:33 |
MSameer | energycsdx: maybe it's time to check gdb | 15:33 |
MSameer | energycsdx: maybe it's time to check with gdb | 15:33 |
*** marxistvegan has quit IRC | 15:54 | |
*** marxistvegan has joined #sailfishos-porters | 15:55 | |
*** PhompAng has quit IRC | 15:59 | |
*** alin has quit IRC | 16:12 | |
*** dhbiker has joined #sailfishos-porters | 16:14 | |
*** Tassadar_ has quit IRC | 16:14 | |
*** carepack has quit IRC | 16:16 | |
*** Tassadar_ has joined #sailfishos-porters | 16:17 | |
*** eleroux has joined #sailfishos-porters | 16:43 | |
sledges | MSameer: energycsdx: yep, let's see that backtrace | 16:52 |
* energycsdx installing debuginfo | 16:52 | |
*** VDVsx has quit IRC | 16:54 | |
*** s5pik3 has joined #sailfishos-porters | 16:58 | |
*** PhompAng has joined #sailfishos-porters | 16:58 | |
*** Tassadar_ has quit IRC | 16:59 | |
*** Tassadar_ has joined #sailfishos-porters | 16:59 | |
*** s5pik3 has quit IRC | 17:01 | |
energycsdx | sledges: it crashed somewhere in /system/lib/liblog.so | 17:04 |
MSameer | get load_sym_files.py from libhybris | 17:07 |
MSameer | source it in gdb | 17:07 |
MSameer | then load-sym-files / | 17:07 |
MSameer | bt | 17:07 |
*** Tassadar_ has quit IRC | 17:09 | |
*** Tassadar_ has joined #sailfishos-porters | 17:09 | |
*** carepack has joined #sailfishos-porters | 17:09 | |
energycsdx | MSameer: it is crashed in strlen called from liblog | 17:13 |
*** Tassadar_ has quit IRC | 17:13 | |
MSameer | sledges: ^ | 17:14 |
*** Tassadar_ has joined #sailfishos-porters | 17:14 | |
energycsdx | i`ll paste logcat | 17:14 |
sledges | hmmm | 17:20 |
sledges | why would it do that, do you have alog/log symlinks ok? never seen.. | 17:21 |
MSameer | might be a null string passed? | 17:21 |
MSameer | null pointer | 17:21 |
*** marxistvegan has quit IRC | 17:25 | |
*** eleroux has quit IRC | 17:25 | |
energycsdx | pastebin.com/sUBwrbUQ | 17:28 |
*** VDVsx has joined #sailfishos-porters | 17:30 | |
*** Tassadar_ has quit IRC | 17:39 | |
*** Tassadar_ has joined #sailfishos-porters | 17:39 | |
*** vrutkovs has quit IRC | 17:41 | |
*** piggz has joined #sailfishos-porters | 17:44 | |
sledges | energycsdx:and where is segfault? im afk atm, this looks like needs deeper digging | 17:58 |
energycsdx | sledges: will look later, i don`t have symbols for android libs atm | 17:59 |
*** PhompAng has quit IRC | 17:59 | |
*** PhompAng has joined #sailfishos-porters | 17:59 | |
*** beidl_ has quit IRC | 18:05 | |
*** beidl has joined #sailfishos-porters | 18:05 | |
*** PhompAng has quit IRC | 18:06 | |
*** piggz has quit IRC | 18:17 | |
*** vrutkovs has joined #sailfishos-porters | 18:23 | |
*** Tassadar_ has quit IRC | 18:24 | |
*** cl has quit IRC | 18:29 | |
*** dhbiker has quit IRC | 18:30 | |
*** carepack has quit IRC | 18:31 | |
*** arcean_ has joined #sailfishos-porters | 18:35 | |
*** arcean has quit IRC | 18:35 | |
*** piggz has joined #sailfishos-porters | 18:37 | |
*** arcean_ is now known as arcean | 18:54 | |
*** furikku has quit IRC | 19:01 | |
*** mostbottom has quit IRC | 19:03 | |
*** oh1jty has quit IRC | 19:03 | |
*** hexo has joined #sailfishos-porters | 19:04 | |
*** piggz has quit IRC | 19:04 | |
*** hexo is now known as mostbottom | 19:05 | |
*** eleroux has joined #sailfishos-porters | 19:09 | |
*** piggz has joined #sailfishos-porters | 19:09 | |
sledges | energycsdx: ok | 19:32 |
*** gogeta_ has joined #sailfishos-porters | 19:44 | |
*** erde74 has joined #sailfishos-porters | 19:47 | |
sledges | energycsdx: this is how it went with nexus5: http://www.merproject.org/logs/%23sailfishos-porters/%23sailfishos-porters.2014-08-04.log.html#t2014-08-04T22:56:06 | 19:47 |
sledges | so if you didn't get "unsupported camera API version" right away, there's still hope to have camera short-term | 19:48 |
sledges | but that was gst1.0 test mind | 19:48 |
sledges | (which you could as well try running) | 19:48 |
*** gogeta_ has quit IRC | 19:50 | |
*** mispp has joined #sailfishos-porters | 19:54 | |
vakkov | Msameer: so BufferQueue( bool allowSynchronousMode = true, int bufferCount = MIN_UNDEQUEUED_BUFFERS); has become BufferQueue(bool allowSynchronousMode = true, const sp<IGraphicBufferAlloc>& allocator = NULL); | 19:58 |
vakkov | in cm10.1; sledges, we used cm10.1 for maguro, right :D | 19:59 |
MSameer | vakkov: you have ANDROID_MAJOR, _MINOR and _MICRO defined | 19:59 |
MSameer | adapt to the new API and use preprocessor directive to control what gets compiled in | 19:59 |
MSameer | I don't know all the android APIs for sure so it's impossible for me to do all the work | 20:00 |
MSameer | the current BufferQueue constructor is something custom I have created and I can change it | 20:01 |
MSameer | which android version is cm10.1 based on? | 20:01 |
vakkov | seems to be 4.2 | 20:02 |
vakkov | 4.2.2 | 20:02 |
*** swex has joined #sailfishos-porters | 20:02 | |
*** SfietKonstantin has joined #sailfishos-porters | 20:08 | |
*** erde74 has quit IRC | 20:09 | |
*** Zotan has joined #sailfishos-porters | 20:19 | |
*** Erikcreed has joined #sailfishos-porters | 20:24 | |
energycsdx | sledges: after load_sym_files.py i have stack trace shorter then without it | 20:40 |
sledges | energycsdx: that may be not important aspect | 20:42 |
energycsdx | https://www.irccloud.com/pastebin/Qcz5AwVM | 20:43 |
*** marxistvegan has joined #sailfishos-porters | 20:44 | |
energycsdx | anyway after i install glibc-debugsource it visible that it tryes print "%s initMorphoPostureDetect failed!!!" | 20:44 |
sledges | yes, which is | 20:45 |
sledges | E/QCameraHWI( 271): int qcamera::MORPHO_FEATURE::initMorphoPostureDetect() | 20:45 |
sledges | or in your case | 20:46 |
sledges | E/QCameraHWI_Preview( 7864): android::status_t android::QCameraStream_preview::initMorphoSceneDetect() : Request nBufSize=9706464, width=1600, height=1200 | 20:46 |
sledges | s/Scene/Posture/ | 20:47 |
sledges | E/QCameraHWI_Preview( 7864): android::status_t android::QCameraStream_preview::initMorphoPostureDetect() : morpho_PostureDetect_start error 80000040 !!! | 20:47 |
sledges | ^ ? | 20:47 |
energycsdx | even don`t know what is MorphoPosture | 20:49 |
sledges | isn't source code at hand) | 20:49 |
sledges | ? | 20:49 |
energycsdx | is it open? | 20:50 |
energycsdx | i can check CAF tree for my device | 20:50 |
sledges | https://android.googlesource.com/platform/hardware/qcom/camera/+/jb-mr1-dev/QCameraHWI_Preview.cpp | 20:51 |
sledges | also it might fail just because droidcamsrc is asking some wrong buffer size | 20:52 |
*** Tassadar has joined #sailfishos-porters | 21:01 | |
*** marxistvegan has quit IRC | 21:02 | |
sledges | energycsdx: how about pastebin GST_DEBUG='droidbufferpool:5' gst-launch-0.10 -e droidcamsrc | 21:02 |
energycsdx | https://www.irccloud.com/pastebin/EZcnYWUi | 21:04 |
sledges | energycsdx: no segfault this time? | 21:05 |
*** arcean has quit IRC | 21:07 | |
energycsdx | https://www.irccloud.com/pastebin/iKPV2YWY | 21:07 |
sledges | thanks, forgot the ! sink | 21:08 |
energycsdx | sledges: looks like it somehow messed with proprietary things like face detection and image adjustments | 21:11 |
energycsdx | https://www.irccloud.com/pastebin/jN3BBKnK | 21:11 |
sledges | we need to simplify the pipeline first | 21:12 |
MSameer | maybe we need to set a certain parameter | 21:12 |
sledges | gst-launch-0.10 -e droidcamsrc ! filesink name="out.raw" | 21:12 |
sledges | haven'ŧ tested yet | 21:12 |
MSameer | if it crashed once then this sounds like a race condition somewhere | 21:13 |
energycsdx | ERROR: from element /GstPipeline:pipeline0/GstFileSink:out.raw: No file name specified for writing. | 21:14 |
MSameer | sledges: it's filename :) | 21:14 |
MSameer | IIRC | 21:14 |
sledges | MSameer: yeap ;) | 21:14 |
sledges | (on the phone now) | 21:14 |
energycsdx | WARNING: erroneous pipeline: no property "filename" in element "filesink0" | 21:15 |
MSameer | energycsdx: try gst-launch-0.10 -e --gst-disable-segtrap droidcamsrc ! droideglsink under gdb | 21:15 |
*** zetaz has joined #sailfishos-porters | 21:17 | |
*** nrta_ has joined #sailfishos-porters | 21:18 | |
energycsdx | https://www.irccloud.com/pastebin/APwQxu0z | 21:21 |
MSameer | sledges: could it be a libhybris unwrapped function? | 21:24 |
*** viper0984 has quit IRC | 21:24 | |
MSameer | energycsdx: could you try the same with the script ? | 21:24 |
sledges | i checked also on jolla phone, and it segfaults before droidbufferpool gstcamerabufferpool.c:273:gst_camera_buffer_pool_set_buffer_count:0x1653088 set buffer count from 0 to 7 (judging from these two: https://www.irccloud.com/pastebin/EZcnYWUi https://www.irccloud.com/pastebin/iKPV2YWY) | 21:24 |
MSameer | sledges: jolla does not segfault | 21:24 |
sledges | precisely, so i used jolla as validator | 21:24 |
sledges | that exactly the same pipeline works fine there | 21:24 |
MSameer | WTF | 21:25 |
MSameer | https://www.irccloud.com/pastebin/iKPV2YWY <- is that jolla? | 21:25 |
sledges | no | 21:25 |
sledges | but it's incomplete pipeline | 21:25 |
sledges | without sink | 21:25 |
sledges | https://www.irccloud.com/pastebin/EZcnYWUi | 21:25 |
sledges | this one ^ | 21:25 |
MSameer | well, I got a crash on jolla today but it was droidcamsrc ! colorconv ! ... | 21:25 |
sledges | jolla does not crash | 21:26 |
sledges | i just thought a pipeline with sink would have different debug trace, so used jolla to validate | 21:26 |
MSameer | ic | 21:26 |
sledges | means that for some reason gst_camera_buffer_pool_set_buffer_count or before would make things segfault | 21:27 |
energycsdx | http://www.techhive.com/article/2029378/morphos-software-will-soon-bring-highend-camera-features-to-smartphones.html | 21:27 |
MSameer | sledges: why do you think that gst_camera_buffer_pool_set_buffer_count crashes? | 21:27 |
energycsdx | morpho is company name | 21:27 |
sledges | MSameer: or it crashes just before it, if we can trust segfault and GST_DEBUG being in sync | 21:28 |
sledges | energycsdx: i see.... | 21:28 |
MSameer | sledges: I only trust gdb | 21:28 |
sledges | energycsdx: would be good to obtain logcat from cyenogen mode there | 21:28 |
sledges | MSameer: fair point | 21:28 |
sledges | energycsdx: when loading android camera | 21:28 |
MSameer | [23:24] <MSameer> energycsdx: could you try the same with the script ? | 21:29 |
MSameer | [23:15] <MSameer> energycsdx: try gst-launch-0.10 -e --gst-disable-segtrap droidcamsrc ! droideglsink under gdb | 21:29 |
MSameer | but use the python script | 21:29 |
MSameer | and don't exit gdb while we are debugging please ;) | 21:29 |
energycsdx | that was with python script | 21:30 |
*** javispedro has joined #sailfishos-porters | 21:30 | |
MSameer | energycsdx: so: frame 0 | 21:30 |
MSameer | info registers | 21:30 |
MSameer | r0 should tell us something | 21:31 |
sledges | MSameer: so this points to Morphos image face detection implants in the stack, possibly missing a libhybris wrap.. | 21:31 |
MSameer | if it's 0x0 then strlen() is getting a NULL pointer | 21:31 |
energycsdx | https://www.irccloud.com/pastebin/tMIU3Bxw | 21:31 |
MSameer | sledges: but we rap vsprintf dont we? | 21:32 |
MSameer | energycsdx: that looks fine... | 21:32 |
MSameer | p ((char *)0x1f371a00)[0] | 21:32 |
energycsdx | this address is not mapped | 21:34 |
energycsdx | https://www.irccloud.com/pastebin/fl9yn70B | 21:34 |
*** SfietKonstantin has quit IRC | 21:34 | |
*** Sfiet_Konstantin has joined #sailfishos-porters | 21:35 | |
MSameer | but that's what r0 carries | 21:35 |
MSameer | odd | 21:35 |
MSameer | energycsdx: any logcat output? | 21:35 |
energycsdx | MSameer: last line is morpho_PostureDetect_start error 80000040 !!! | 21:37 |
MSameer | a few ones before that please? | 21:37 |
energycsdx | https://www.irccloud.com/pastebin/xIAT6x0u | 21:38 |
MSameer | looks like we will have to strace it :/ | 21:39 |
energycsdx | https://www.irccloud.com/pastebin/juXVnXGU | 21:43 |
energycsdx | looks like it need to decompile that morpho thing | 21:43 |
MSameer | i need like 200 lines before that :D | 21:43 |
MSameer | with strace -f | 21:44 |
MSameer | mmmmm | 21:45 |
MSameer | wait a second | 21:45 |
MSameer | you said that camera app starts | 21:45 |
MSameer | but it gives you a blank preview | 21:45 |
MSameer | so the crash does not happen with camera app | 21:45 |
MSameer | maybe it does that because it does not even reach that stage | 21:47 |
sledges | or it happens in it's gst pipeline equiv thread which doesn't take the app out? | 21:47 |
energycsdx | http://pastebin.com/S07xD12q | 21:48 |
MSameer | sledges: nop. a crash in a thread has the same effect | 21:48 |
sledges | ook | 21:48 |
MSameer | we don't mask any signals AFAICT | 21:48 |
sledges | energycsdx: strace -f pls :) | 21:48 |
sledges | sorry for being a pain:) | 21:49 |
MSameer | energycsdx: :D | 21:49 |
MSameer | i wonder if sony needs some extra parameters | 21:51 |
*** marxistvegan has joined #sailfishos-porters | 21:51 | |
MSameer | wait4(-1, NULL, 0, NULL) = -1 ECHILD (No child processes) | 21:52 |
MSameer | not related | 21:52 |
MSameer | sledges: I think it's obvious. It tries to access an unmapped area for whatever reason | 21:53 |
MSameer | --- SIGCHLD {si_signo=SIGCHLD, si_code=CLD_EXITED, si_pid=2620, si_status=0, si_utime=0, si_stime=3} --- | 21:53 |
MSameer | err | 21:53 |
MSameer | --- SIGSEGV {si_signo=SIGSEGV, si_code=SEGV_MAPERR, si_addr=0xd48e9600} --- | 21:53 |
MSameer | energycsdx: can you try running the same pipeline as root? | 21:54 |
energycsdx | do you know pastebin with limit more then 512k? | 21:55 |
MSameer | energycsdx: dropbox ? ;) | 21:55 |
energycsdx | https://drive.google.com/file/d/0B7-IjLFjJHzjc0MyZ1ZoMHM2U0E/view?usp=sharing | 21:58 |
MSameer | sledges: ^ | 21:59 |
MSameer | sledges: it's really obvious. a crash while logging? | 22:00 |
MSameer | maybe there is a function which we don't wrap? | 22:01 |
MSameer | 2836 writev(5, [{"\6", 1}, {"QCameraHWI_Preview\0", 19}, {"android::status_t android::QCame"..., 124}], 3 <unfinished ...> | 22:01 |
MSameer | 2836 <... writev resumed> ) = 144 | 22:01 |
MSameer | 2836 --- SIGSEGV {si_signo=SIGSEGV, si_code=SEGV_MAPERR, si_addr=0x36780b00} --- | 22:01 |
MSameer | I don't know what could that function be | 22:01 |
sledges | no clues here: https://www.irccloud.com/pastebin/APwQxu0z ? | 22:04 |
MSameer | vsnprintf ? | 22:05 |
sledges | it's glibc | 22:06 |
sledges | http://developerweb.net/viewtopic.php?pid=31490#p31490 | 22:06 |
MSameer | sledges: I don't get you | 22:07 |
sledges | it's not a bionic function, so no need to consider a wrapper one | 22:07 |
sledges | as the comment says, it might be dynamic heap corruption | 22:07 |
sledges | wonder if valgrind could spot in our env | 22:08 |
sledges | *hybris env | 22:08 |
MSameer | I guess it should | 22:08 |
MSameer | energycsdx: could you try under valgrind? | 22:08 |
MSameer | sledges: how do you know it's glibc? | 22:08 |
sledges | http://osxr.org/glibc/ident?_i=_IO_vfprintf_internal | 22:08 |
MSameer | sledges: I feel my next question will get google.com as an answer :P | 22:11 |
sledges | :D | 22:11 |
sledges | i still wonder if it's segfault while logging, where & why does this pop up: E/QCameraHWI_Preview( 2263): android::status_t android::QCameraStream_preview::initMorphoPostureDetect() : morpho_PostureDetect_start error 80000040 !!! | 22:12 |
MSameer | sledges: but vsnprintf is not in hybris hooks | 22:12 |
MSameer | #2 0x40333674 in _IO_vsnprintf (string=0xbefff004 "", maxlen=<optimized out>, format=0x40a76747 "%s initMorphoPostureDetect failed!!!", args=...) | 22:13 |
MSameer | initMorphoPostureDetect failed!!! <- this is what crashes? | 22:13 |
sledges | yes | 22:14 |
sledges | lol | 22:14 |
sledges | it tries to inform us about failure | 22:14 |
sledges | and fails while trying | 22:14 |
sledges | :D | 22:14 |
*** Sfiet_Konstantin has quit IRC | 22:14 | |
MSameer | sledges: but vsnprintf is not in hybris hooks | 22:15 |
sledges | so you reckon it just calls a bionic one and then blows up, this what happens? | 22:16 |
MSameer | you are the bionic expert :D | 22:16 |
sledges | im glad im not :D | 22:16 |
MSameer | http://osxr.org/android/ident?_i=android_openEventTagMap | 22:18 |
sledges | lol 0348 fprintf(stderr, "ARRRRGH\n"); | 22:21 |
MSameer | ? | 22:22 |
MSameer | heh | 22:23 |
MSameer | android :p | 22:23 |
sledges | just browzing that android_tagMap thing | 22:23 |
sledges | something doesn't compute: where would android_openEventTagMap call (vsn)printf with "%s initMorphoPostureDetect failed!!!" string? | 22:23 |
sledges | https://www.irccloud.com/pastebin/APwQxu0z | 22:24 |
sledges | another thread? | 22:24 |
MSameer | looks like a thread gets created to iitialize that thingy and it fails | 22:25 |
MSameer | so the original thread outputs that error? | 22:25 |
MSameer | I can't tell for sure | 22:25 |
MSameer | we need Richard to read the assembly and tell us :D | 22:25 |
sledges | and erm | 22:26 |
*** eleroux has quit IRC | 22:26 | |
sledges | oh | 22:26 |
sledges | hang on | 22:26 |
sledges | MSameer: https://github.com/libhybris/libhybris/commit/26bf55a60677451547ef88a3ace379b6df029493#diff-daf552d42577386767c912568c97e666L1347 | 22:28 |
sledges | is all hooked | 22:28 |
sledges | is just dynamic now | 22:28 |
*** Erikcreed has quit IRC | 22:38 | |
MSameer | sledges: i will trust you | 22:43 |
MSameer | energycsdx: any progress? | 22:43 |
*** zetaz has left #sailfishos-porters | 22:43 | |
sledges | MSameer: if that's not in, all other deleted functions wouldn't be either;) | 22:44 |
*** Tassadar has quit IRC | 22:51 | |
energycsdx | MSameer: decompiling | 22:59 |
MSameer | energycsdx: no need. just run valgrind please | 23:00 |
*** piggz has quit IRC | 23:02 | |
energycsdx | MSameer: http://pastebin.com/aBgcggKf | 23:07 |
MSameer | sledges: | 23:09 |
MSameer | ==3541== Invalid read of size 1 | 23:09 |
MSameer | ==3541== at 0x483992C: strlen (mc_replace_strmem.c:404) | 23:09 |
MSameer | ==3541== by 0x4B2CBC7: vfprintf (vfprintf.c:1987) | 23:09 |
MSameer | ==3541== by 0x4B4F673: vsnprintf (vsnprintf.c:120) | 23:09 |
MSameer | ==3541== by 0x547F3DD: __android_log_print (in /system/lib/liblog.so) | 23:09 |
MSameer | ==3541== Address 0xb8d15100 is not stack'd, malloc'd or (recently) free'd | 23:09 |
sledges | exactly what i was looking at :) | 23:09 |
*** vakkov has quit IRC | 23:09 | |
sledges | it should be using liblog from ours | 23:09 |
MSameer | that's what I said :/ | 23:09 |
sledges | well said MSameer ;) | 23:09 |
MSameer | crash while print | 23:09 |
sledges | energycsdx: ls -l /usr/libexec/droid-hybris/lib/liblog* | 23:09 |
MSameer | sledges: I really have no idea how to fix that | 23:10 |
sledges | wrong indirection, possibly staticly linked | 23:10 |
energycsdx | sledges: whole /system is built with hybris patch | 23:11 |
sledges | hmm is not how we play ;) | 23:11 |
energycsdx | i don`t mount device /system partition | 23:11 |
sledges | so if /system/lib/liblog.so goes to /dev/log and 100% not /dev/alog, means someting else goes wrong | 23:12 |
MSameer | it looks like an invalid pointer being passed around? | 23:16 |
sledges | from Morphos world? | 23:16 |
sledges | energycsdx: the other way round - /dev/alog should be the point of choice (https://github.com/mer-hybris/android_system_core/commit/c5b5689edd8aeba08f0718450766a8d3fcb92bc8) | 23:16 |
sledges | i always mix the two:P now i'll know why | 23:16 |
MSameer | alog is ours | 23:18 |
energycsdx | that morpho lib calls syscall function, which is strange | 23:19 |
sledges | closed bastards can do anything the like.. | 23:19 |
energycsdx | there is ~20 functions that lib imports | 23:20 |
energycsdx | mostly pthreads | 23:20 |
sledges | could you gut it out altogether? | 23:20 |
sledges | provide a dummy if inevitable? | 23:20 |
*** javispedro has quit IRC | 23:21 | |
energycsdx | get? | 23:22 |
sledges | gut, cut, unroot | 23:22 |
energycsdx | https://www.irccloud.com/pastebin/w3UAvuSU | 23:23 |
sledges | i mean | 23:24 |
MSameer | abort :p | 23:24 |
sledges | try to disable Morphos lib altogether | 23:25 |
sledges | replace it with foobar(){return 0;} one for every function | 23:25 |
MSameer | sledges: wasn't it that we cannot hook into bionic <-> bionic calls with hybris? | 23:25 |
MSameer | energycsdx: maybe just move it away for now | 23:25 |
sledges | yea, like the .so file :)) | 23:26 |
sledges | decouple:) | 23:26 |
energycsdx | sledges: what if i build hardware/qcom/camera from CAF tree? | 23:27 |
sledges | MSameer: yep i think there's nothing we can intervene between bionic-bionic | 23:27 |
sledges | energycsdx: also an option | 23:27 |
energycsdx | sledges: how it is on sbj? | 23:28 |
sledges | well, Morphos is the first time i hear abouit | 23:28 |
*** mispp has quit IRC | 23:31 | |
energycsdx | sledges: i mean, is camera bits compiled from source or odm just provides blobs? | 23:31 |
sledges | don't know that | 23:31 |
*** iTune has joined #sailfishos-porters | 23:32 | |
MSameer | energycsdx: might be beneficial to get cm logcat too | 23:35 |
MSameer | sledges: but why can't we do anything for bionic-bionic? | 23:35 |
sledges | MSameer: totally unaware about that;) | 23:36 |
MSameer | ok | 23:36 |
MSameer | fair enough | 23:36 |
*** vakkov has joined #sailfishos-porters | 23:37 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!