Login at Kodi Home

Koying · 2014-11-19, 17:45

For ios, I guess the inodes are always < 2^32, then.
There is no significant truncation and no issue in this case.

gdhillon · 2014-11-19, 17:45

@Koying old version still crashing for me.

Koying · (This post was last modified: 2014-11-19, 17:59 by Koying.)

Ok. Thanks. Will make this into a proper PR.
I doubt it will be still accepted for mainstream Kodi, though, so you'll have to rely on special builds for a while.

[EDIT] Or Google implements a workaround in a future firmware, of course, but they will probably patch their own sqlite, which won't solve our own issue.

jelte · 2014-11-19, 18:00

Thanks for this testbuild, works great. Even with external databases and media.

Jorg van tiel · 2014-11-19, 20:11

Thanks Kody. The test build works perfect on the N9

gdhillon · 2014-11-20, 03:35

@Koying thanks again, let us know when it get merged with main build.

**Memphiz** · 2014-11-20, 16:18

Kodi + Koying == Kody ... muhahahaa thats an idea.

Koying · 2014-11-20, 17:27

I have difficulties finding out why the installation fails on the neon check.
Could you test http://mirrors.xbmc.org/test-builds/andr...bi-v7a.apk and provide a logcat, please.

karhukuoma · 2014-11-20, 20:35

(2014-11-20, 17:27)Koying Wrote:

I have difficulties finding out why the installation fails on the neon check.
Could you test http://mirrors.xbmc.org/test-builds/andr...bi-v7a.apk and provide a logcat, please.

http://pastebin.com/4kByxeQ1

Koying · 2014-11-20, 21:11

Thanks. Apparently, what is returned to Kodi is not the same as what reported via adb. I've read a mention of that somewhere.
Anyway, aarch64 is neon enabled by definition, I read.

Quote:D/Kodi ( 2091): /proc/cpuinfo = Processor : NVIDIA Denver 1.0 rev 0 (aarch64)
D/Kodi ( 2091): processor : 0
D/Kodi ( 2091): processor : 1
D/Kodi ( 2091): Features : fp asimd aes pmull sha1 sha2 crc32
D/Kodi ( 2091): CPU implementer : 0x4e
D/Kodi ( 2091): CPU architecture: AArch64
D/Kodi ( 2091): CPU variant : 0x0
D/Kodi ( 2091): CPU part : 0x000
D/Kodi ( 2091): CPU revision : 0
D/Kodi ( 2091):
D/Kodi ( 2091): Hardware : Flounder
D/Kodi ( 2091): Revision : 0000
D/Kodi ( 2091): Serial : 0000000000000000

Koying · 2014-11-20, 23:53

https://github.com/xbmc/xbmc/pull/5773

Could you test http://mirrors.xbmc.org/test-builds/andr...bi-v7a.apk to confirm it's ok, please.

karhukuoma · (This post was last modified: 2014-11-21, 00:40 by karhukuoma.)

(2014-11-20, 23:53)Koying Wrote: https://github.com/xbmc/xbmc/pull/5773

Could you test http://mirrors.xbmc.org/test-builds/andr...bi-v7a.apk to confirm it's ok, please.

Installs, but crashes when trying to play videos atleast from SMB or UPnP sources. Local file seems to work. I can get the logcat this evening if no one gets it first.

**hurlbert** · 2014-11-21, 02:41

(2014-11-21, 00:22)karhukuoma Wrote:
(2014-11-20, 23:53)Koying Wrote: https://github.com/xbmc/xbmc/pull/5773

Could you test http://mirrors.xbmc.org/test-builds/andr...bi-v7a.apk to confirm it's ok, please.

Installs, but crashes when trying to play videos atleast from SMB or UPnP sources. Local file seems to work. I can get the logcat this evening if no one gets it first.

Just FYI - I got an OTA on the Nexus 9 earlier this afternoon. I installed it.

I went and pulled this APK and installed it and it did indeed play local mp4 and divx perfectly.

I setup the smb to my NAS and the it played divx and mpg files from there as well.

I've only played with it for a few minutes but so far it's been perfect.

Android version: 5.0

Kernel version: 3.10.40-ga3846f1
Thu Nov 6 01:39:45 UTC 2014

Build Number: LRX21R

Thank you so much. You should feel bad because now I can go workout while watching my dvr'd game! Damn you!!!

karhukuoma · 2014-11-21, 16:55

(2014-11-21, 00:22)karhukuoma Wrote:
(2014-11-20, 23:53)Koying Wrote: https://github.com/xbmc/xbmc/pull/5773

Could you test http://mirrors.xbmc.org/test-builds/andr...bi-v7a.apk to confirm it's ok, please.

Installs, but crashes when trying to play videos atleast from SMB or UPnP sources. Local file seems to work. I can get the logcat this evening if no one gets it first.

Does this seem to be releted to nexus9? Same upnp source and file works on nexus5 lollipop with xbmc gotham 13.1.

Code:
V/XBMC    ( 6245): Debug Print: void XCURL::DllLibCurlGlobal::easy_aquire(const char*, const char*, XCURL::CURL_HANDLE**, XCURL::CURLM**) - Created session to http://192.168.0.2

V/XBMC    ( 6245): Debug Print: ------ Window Init (DialogBusy.xml) ------

V/XBMC    ( 6245): Debug Print: Window DialogBusy.xml was already loaded

V/XBMC    ( 6245): Debug Print: Alloc resources: 0.01m

V/XBMC    ( 6245): Debug Print: Unfocus WindowID: 10025, ControlID: 50

V/XBMC    ( 6245): Debug Print: SECTION:UnloadDelayed(DLL: libImageLib-arm.so)

V/XBMC    ( 6245): Debug Print: Unloading: /libImageLib-arm.so

V/XBMC    ( 6245): Debug Print: CCurlFile::GetMimeType - http://192.168.0.2:5001/get/0$1$11/south.park.1806.mp4 -> video/mpeg

V/XBMC    ( 6245): Debug Print: CFileCache::Open - opening <get/0$1$11/south.park.1806.mp4> using cache

V/XBMC    ( 6245): Debug Print: CurlFile::Open(0xac2d8cf8) http://192.168.0.2:5001/get/0$1$11/south.park.1806.mp4

V/XBMC    ( 6245): Debug Print: Creating Demuxer

V/XBMC    ( 6245): Debug Print: Thread FileCache start, auto delete: false

V/XBMC    ( 6245): Debug Print: bool CDVDDemuxFFmpeg::Open(CDVDInputStream*, bool, bool) - probing detected format [mpegts]

V/XBMC    ( 6245): Debug Print: bool CDVDDemuxFFmpeg::Open(CDVDInputStream*, bool, bool) - avformat_find_stream_info starting

V/XBMC    ( 6245): Debug Print: bool CDVDDemuxFFmpeg::Open(CDVDInputStream*, bool, bool) - av_find_stream_info finished

V/XBMC    ( 6245): Debug Print: Opening stream: 0 source: 256

V/XBMC    ( 6245): Debug Print: Creating video codec with codec id: 28

V/XBMC    ( 6245): Debug Print: CDVDFactoryCodec: compiled in hardware support: AMCodec:yes MediaCodec:yes OpenMax:no libstagefright:yes VDPAU:no VAAPI:no iMXVPU:no

V/XBMC    ( 6245): Debug Print: MediaCodec Video Decoder...

V/XBMC    ( 6245): Debug Print: FactoryCodec - Video: mediacodec - Opening

W/VideoCapabilities( 6245): Unsupported mime video/mpeg2

W/AudioCapabilities( 6245): Unsupported mime audio/mpeg-L2

W/VideoCapabilities( 6245): Unrecognized level 0 for video/x-vnd.on2.vp8

I/VideoCapabilities( 6245): Unsupported profile 4 for video/mp4v-es

I/OMXClient( 6245): Using client-side OMX mux.

V/XBMC    ( 6245): Debug Print: CDVDVideoCodecAndroidMediaCodec::Open m_codecname(OMX.Nvidia.h264.decode), colorFormat(2135033992)

V/XBMC    ( 6245): Debug Print: CDVDVideoCodecAndroidMediaCodec::Open m_codecname(OMX.Nvidia.h264.decode), colorFormat(21)

W/ACodec  ( 6245): cannot describe color format 0x15 = 21 with stride=0 and sliceHeight=404

F/ACodec  ( 6245): frameworks/av/media/libstagefright/ACodec.cpp:3232 CHECK_LE( rect.nLeft + rect.nWidth - 1,videoDef->nFrameWidth) failed: 4294967295 vs. 0

--------- beginning of crash

F/libc    ( 6245): Fatal signal 6 (SIGABRT), code -6 in tid 6429 (CodecLooper)

I/libc    ( 6245): Suppressing debuggerd output because prctl(PR_GET_DUMPABLE)==0

I/WindowState(18195): WIN DEATH: Window{35d0158c u0 org.xbmc.kodi/org.xbmc.kodi.Main}

E/OMXNodeInstance(17949): !!! Observer died. Quickly, do something, ... anything...

I/Zygote  ( 1077): Process 6245 exited due to signal (6)

I/ActivityManager(18195): Process org.xbmc.kodi (pid 6245) has died

W/ActivityManager(18195): Force removing ActivityRecord{361af9eb u0 org.xbmc.kodi/.Main t821}: app died, no saved state

W/SearchServiceClient(19365): Attempt to handover from detached client

Koying · 2014-11-21, 17:18

N9 is asserting internally because it doesn't like your file (pretty stupid of them).
Try to disable mediacodec.