Advent Vega kernel source code now available!

[Guest ejtagle]

As of yet i cannot play any video. I get this error in logcat:

20:46:54.279 Error WVMExtractor 103  Failed to open libwvm.so

I guess it has to do with the prop.files.

Don't worry about that error. libwvm.so is a propietary lib used to demux drm content. That is not the cause of the problem :)

[Guest Scanno]

Don't worry about that error. libwvm.so is a propietary lib used to demux drm content. That is not the cause of the problem :)

It is stopping me watching Youtube videos

Audio is working... Am listening to music through headphones

Some logcat:

21:36:48.176 Info AwesomePlayer 103  setDataSource_l('http://redirector.c.youtube.com/videoplayback?id=5efffee9743222ad&itag=22&source=youtube&uaopt=no-save&upn=YVFzsvB1Nlg&el=videos&devKey=ATpxuMO4AN5NR1nGmfaQDBQO88HsQjpE1a8d1GxQnGDm&app=youtube_gdata&ip=0.0.0.0&ipbits=0&expire=1342755390&sparams=id,itag,source,uaopt,upn,ip,ipbits,expire&signature=1DDAABF6DD68E135B7E3F595C7A96A49DC7CC32D.21B05667B6CDB41CD3D37C569DDA456744D0CD&key=yta1&androidcid=mvapp-unknown&dnc=1')

21:36:48.176 Verbose ChromiumHTTPDataSource 103  connect on behalf of uid 10058

21:36:48.176 Info ChromiumHTTPDataSource 103  connect to http://redirector.c.youtube.com/videoplayback?id=5efffee9743222ad&itag=22&source=youtube&uaopt=no-save&upn=YVFzsvB1Nlg&el=videos&devKey=ATpxuMO4AN5NR1nGmfaQDBQO88HsQjpE1a8d1GxQnGDm&app=youtube_gdata&ip=0.0.0.0&ipbits=0&expire=1342755390&sparams=id,itag,source,uaopt,upn,ip,ipbits,expire&signature=1DDAABF6DD68E135B7E3F595C7A96A49DC7CC32D.21B05667B6CDB41CD3D37C569DDA456744D0CD&key=yta1&androidcid=mvapp-unknown&dnc=1 @0

21:36:48.262 Info ActivityManager 271  Displayed com.google.android.youtube/.app.honeycomb.tablet.WatchActivity: +1s135ms

21:36:48.505 Error WVMExtractor 103  Failed to open libwvm.so

21:36:48.687 Debug dalvikvm 2552  GC_CONCURRENT freed 2174K, 11% free 21345K/23879K, paused 5ms+15ms, total 209ms

21:36:48.708 Debug dalvikvm 2552  WAIT_FOR_CONCURRENT_GC blocked 51ms

21:36:48.708 Debug dalvikvm 2552  WAIT_FOR_CONCURRENT_GC blocked 27ms

21:36:48.729 Info OMXCodec 103  [OMX.google.h264.decoder] AVC profile = 100 (High), level = 31

21:36:48.729 Info OMXCodec 103  [OMX.google.h264.decoder] video dimensions are 320 x 240

21:36:48.769 Info OMXCodec 103  [OMX.google.h264.decoder] Crop rect is 320 x 240 @ (0, 0)

21:36:49.505 Debug dalvikvm 2552  GC_CONCURRENT freed 1634K, 9% free 21759K/23879K, paused 20ms+12ms, total 163ms

21:36:49.558 Debug dalvikvm 2552  WAIT_FOR_CONCURRENT_GC blocked 68ms

21:36:49.558 Debug dalvikvm 2552  WAIT_FOR_CONCURRENT_GC blocked 74ms

21:36:49.558 Debug dalvikvm 2552  WAIT_FOR_CONCURRENT_GC blocked 67ms

21:36:49.871 Error SoftAVC 103  Decoder failed: -2

21:36:49.889 Error OMXCodec 103  [OMX.google.h264.decoder] ERROR(0x80001001, -1007)

21:36:49.889 Info SoftAAC2 103  Reconfiguring decoder: 44100 Hz, 2 channels

21:36:49.889 Error MediaPlayer 2552  error (1, -2147483648)

21:36:49.950 Debug dalvikvm 271  GC_CONCURRENT freed 1116K, 19% free 10466K/12871K, paused 22ms+10ms, total 135ms

21:36:50.002 Debug dalvikvm 2552  GC_FOR_ALLOC freed 1610K, 10% free 21858K/24199K, paused 72ms, total 72ms

21:36:50.028 Error MediaPlayer 2552  Error (1,-2147483648)

21:36:50.028 Warning YouTube 2552  MediaPlayer error during playback [what=1, extra=-2147483648]

21:36:50.028 Warning YouTube 2552  Retrying MediaPlayer error [retry=1, max=3]

21:36:50.028 Warning TimedEventQueue 103  Event 3 was not found in the queue, already cancelled?

21:36:50.028 Warning AudioFlinger 103  session id 128 not found for pid 2552

21:36:50.028 Debug MediaPlayer 2552  Couldn't open file on client side, trying server side

21:36:50.028 Info AwesomePlayer 103  setDataSource_l('http://redirector.c.youtube.com/videoplayback?id=5efffee9743222ad&itag=22&source=youtube&uaopt=no-save&upn=YVFzsvB1Nlg&el=videos&devKey=ATpxuMO4AN5NR1nGmfaQDBQO88HsQjpE1a8d1GxQnGDm&app=youtube_gdata&ip=0.0.0.0&ipbits=0&expire=1342755390&sparams=id,itag,source,uaopt,upn,ip,ipbits,expire&signature=1DDAABF6DD68E135B7E3F595C7A96A49DC7CC32D.21B05667B6CDB41CD3D37C569DDA456744D0CD&key=yta1&androidcid=mvapp-unknown&dnc=1')

21:36:50.028 Verbose ChromiumHTTPDataSource 103  connect on behalf of uid 10058

21:36:50.028 Info ChromiumHTTPDataSource 103  connect to http://redirector.c.youtube.com/videoplayback?id=5efffee9743222ad&itag=22&source=youtube&uaopt=no-save&upn=YVFzsvB1Nlg&el=videos&devKey=ATpxuMO4AN5NR1nGmfaQDBQO88HsQjpE1a8d1GxQnGDm&app=youtube_gdata&ip=0.0.0.0&ipbits=0&expire=1342755390&sparams=id,itag,source,uaopt,upn,ip,ipbits,expire&signature=1DDAABF6DD68E135B7E3F595C7A96A49DC7CC32D.21B05667B6CDB41CD3D37C569DDA456744D0CD&key=yta1&androidcid=mvapp-unknown&dnc=1 @0

21:36:50.305 Error WVMExtractor 103  Failed to open libwvm.so

21:36:50.512 Info OMXCodec 103  [OMX.google.h264.decoder] AVC profile = 100 (High), level = 31

21:36:50.512 Info OMXCodec 103  [OMX.google.h264.decoder] video dimensions are 320 x 240

21:36:50.555 Info OMXCodec 103  [OMX.google.h264.decoder] Crop rect is 320 x 240 @ (0, 0)

21:36:51.534 Error SoftAVC 103  Decoder failed: -2

21:36:51.584 Error OMXCodec 103  [OMX.google.h264.decoder] ERROR(0x80001001, -1007)

21:36:51.584 Info SoftAAC2 103  Reconfiguring decoder: 44100 Hz, 2 channels

21:36:51.584 Error MediaPlayer 2552  error (1, -2147483648)

21:36:51.584 Error MediaPlayer 2552  Error (1,-2147483648)

21:36:51.584 Warning YouTube 2552  MediaPlayer error during playback [what=1, extra=-2147483648]

21:36:51.584 Warning YouTube 2552  Retrying MediaPlayer error [retry=2, max=3]

21:36:51.584 Warning TimedEventQueue 103  Event 3 was not found in the queue, already cancelled?

21:36:51.584 Warning AudioFlinger 103  session id 129 not found for pid 2552

21:36:51.584 Debug MediaPlayer 2552  Couldn't open file on client side, trying server side

21:36:51.584 Info AwesomePlayer 103  setDataSource_l('http://redirector.c.youtube.com/videoplayback?id=5efffee9743222ad&itag=22&source=youtube&uaopt=no-save&upn=YVFzsvB1Nlg&el=videos&devKey=ATpxuMO4AN5NR1nGmfaQDBQO88HsQjpE1a8d1GxQnGDm&app=youtube_gdata&ip=0.0.0.0&ipbits=0&expire=1342755390&sparams=id,itag,source,uaopt,upn,ip,ipbits,expire&signature=1DDAABF6DD68E135B7E3F595C7A96A49DC7CC32D.21B05667B6CDB41CD3D37C569DDA456744D0CD&key=yta1&androidcid=mvapp-unknown&dnc=1')

21:36:51.584 Verbose ChromiumHTTPDataSource 103  connect on behalf of uid 10058

21:36:51.584 Info ChromiumHTTPDataSource 103  connect to http://redirector.c.youtube.com/videoplayback?id=5efffee9743222ad&itag=22&source=youtube&uaopt=no-save&upn=YVFzsvB1Nlg&el=videos&devKey=ATpxuMO4AN5NR1nGmfaQDBQO88HsQjpE1a8d1GxQnGDm&app=youtube_gdata&ip=0.0.0.0&ipbits=0&expire=1342755390&sparams=id,itag,source,uaopt,upn,ip,ipbits,expire&signature=1DDAABF6DD68E135B7E3F595C7A96A49DC7CC32D.21B05667B6CDB41CD3D37C569DDA456744D0CD&key=yta1&androidcid=mvapp-unknown&dnc=1 @0

21:36:51.872 Error WVMExtractor 103  Failed to open libwvm.so

21:36:52.076 Info OMXCodec 103  [OMX.google.h264.decoder] AVC profile = 100 (High), level = 31

21:36:52.076 Info OMXCodec 103  [OMX.google.h264.decoder] video dimensions are 320 x 240

21:36:52.104 Info OMXCodec 103  [OMX.google.h264.decoder] Crop rect is 320 x 240 @ (0, 0)

21:36:53.120 Error SoftAVC 103  Decoder failed: -2

21:36:53.147 Error OMXCodec 103  [OMX.google.h264.decoder] ERROR(0x80001001, -1007)

21:36:53.147 Info SoftAAC2 103  Reconfiguring decoder: 44100 Hz, 2 channels

21:36:53.147 Error MediaPlayer 2552  error (1, -2147483648)

21:36:53.147 Error MediaPlayer 2552  Error (1,-2147483648)

21:36:53.148 Warning YouTube 2552  MediaPlayer error during playback [what=1, extra=-2147483648]

21:36:53.148 Warning YouTube 2552  Retrying MediaPlayer error [retry=3, max=3]

21:36:53.148 Warning TimedEventQueue 103  Event 3 was not found in the queue, already cancelled?

21:36:53.148 Warning AudioFlinger 103  session id 130 not found for pid 2552

21:36:53.148 Debug MediaPlayer 2552  Couldn't open file on client side, trying server side

21:36:53.148 Info AwesomePlayer 103  setDataSource_l('http://redirector.c.youtube.com/videoplayback?id=5efffee9743222ad&itag=22&source=youtube&uaopt=no-save&upn=YVFzsvB1Nlg&el=videos&devKey=ATpxuMO4AN5NR1nGmfaQDBQO88HsQjpE1a8d1GxQnGDm&app=youtube_gdata&ip=0.0.0.0&ipbits=0&expire=1342755390&sparams=id,itag,source,uaopt,upn,ip,ipbits,expire&signature=1DDAABF6DD68E135B7E3F595C7A96A49DC7CC32D.21B05667B6CDB41CD3D37C569DDA456744D0CD&key=yta1&androidcid=mvapp-unknown&dnc=1')

21:36:53.148 Verbose ChromiumHTTPDataSource 103  connect on behalf of uid 10058

21:36:53.148 Info ChromiumHTTPDataSource 103  connect to http://redirector.c.youtube.com/videoplayback?id=5efffee9743222ad&itag=22&source=youtube&uaopt=no-save&upn=YVFzsvB1Nlg&el=videos&devKey=ATpxuMO4AN5NR1nGmfaQDBQO88HsQjpE1a8d1GxQnGDm&app=youtube_gdata&ip=0.0.0.0&ipbits=0&expire=1342755390&sparams=id,itag,source,uaopt,upn,ip,ipbits,expire&signature=1DDAABF6DD68E135B7E3F595C7A96A49DC7CC32D.21B05667B6CDB41CD3D37C569DDA456744D0CD&key=yta1&androidcid=mvapp-unknown&dnc=1 @0

21:36:53.446 Error WVMExtractor 103  Failed to open libwvm.so

21:36:53.653 Info OMXCodec 103  [OMX.google.h264.decoder] AVC profile = 100 (High), level = 31

21:36:53.699 Info OMXCodec 103  [OMX.google.h264.decoder] video dimensions are 320 x 240

21:36:53.699 Info OMXCodec 103  [OMX.google.h264.decoder] Crop rect is 320 x 240 @ (0, 0)

21:36:55.676 Error SoftAVC 103  Decoder failed: -2

21:36:55.704 Error OMXCodec 103  [OMX.google.h264.decoder] ERROR(0x80001001, -1007)

21:36:55.704 Info SoftAAC2 103  Reconfiguring decoder: 44100 Hz, 2 channels

21:36:55.704 Error MediaPlayer 2552  error (1, -2147483648)

21:36:55.704 Debug audio_hw_primary 103  start_output_stream: device:0, rate:44100

21:36:55.738 Warning audio_hw_primary 103  out_write() limiting sleep time 79727 to 39909

21:36:55.738 Error MediaPlayer 2552  Error (1,-2147483648)

21:36:55.738 Warning YouTube 2552  MediaPlayer error during playback [what=1, extra=-2147483648]

21:36:55.738 Warning YouTube 2552  Reporting MediaPlayer error

21:36:55.738 Warning TimedEventQueue 103  Event 4 was not found in the queue, already cancelled?

21:36:55.738 Warning AudioFlinger 103  session id 131 not found for pid 2552

21:36:55.752 Warning audio_hw_primary 103  out_write() limiting sleep time 59591 to 39909

EDIT:

Hmmm am i missing something? I am using the newer prop. files you linked to (or was it brucelee666) a while back.

Edited July 19, 2012 by Scanno

[Guest Scanno]

Eduardo,

Just got this by playing video. Perhaps it says something to you:

21:42:27.347 Debug dalvikvm 2797  WAIT_FOR_CONCURRENT_GC blocked 51ms

21:42:28.073 Info OMXCodec 103  [OMX.google.h264.decoder] video dimensions are 480 x 368

21:42:28.103 Info OMXCodec 103  [OMX.google.h264.decoder] Crop rect is 480 x 360 @ (0, 0)

21:42:28.103 Info OMXCodec 103  [OMX.google.h264.decoder] video dimensions are 480 x 368

21:42:28.103 Info OMXCodec 103  [OMX.google.h264.decoder] Crop rect is 480 x 360 @ (0, 0)

21:42:28.103 Info SoftAAC2 103  Reconfiguring decoder: 44100 Hz, 2 channels

21:42:28.125 Debug audio_hw_primary 103  start_output_stream: device:0, rate:44100

21:42:29.635 Debug dalvikvm 271  GC_CONCURRENT freed 1069K, 24% free 10470K/13703K, paused 2ms+7ms, total 53ms

21:42:31.803 Debug dalvikvm 2797  GC_CONCURRENT freed 1954K, 11% free 19899K/22215K, paused 13ms+8ms, total 90ms

21:42:31.858 Debug dalvikvm 2797  WAIT_FOR_CONCURRENT_GC blocked 65ms

21:42:36.921 Debug dalvikvm 2797  GC_FOR_ALLOC freed 1365K, 12% free 19872K/22343K, paused 54ms, total 54ms

21:42:42.487 Debug dalvikvm 2797  --- called into dvmAbstractMethodStub

21:42:42.487 Debug AndroidRuntime 2797  Shutting down VM

21:42:42.514 Warning dalvikvm 2797  threadid=1: thread exiting with uncaught exception (group=0x41fa9300)

21:42:42.514 Error AndroidRuntime 2797  FATAL EXCEPTION: main

21:42:42.514 Error AndroidRuntime 2797  java.lang.AbstractMethodError: abstract method not implemented

21:42:42.514 Error AndroidRuntime 2797   at android.media.MediaRouter$Callback.onRouteVolumeChanged(MediaRouter.java)

21:42:42.514 Error AndroidRuntime 2797   at android.media.MediaRouter.dispatchRouteVolumeChanged(MediaRouter.java:592)

21:42:42.514 Error AndroidRuntime 2797   at android.media.MediaRouter.systemVolumeChanged(MediaRouter.java:603)

21:42:42.514 Error AndroidRuntime 2797   at android.media.MediaRouter$VolumeChangeReceiver.onReceive(MediaRouter.java:1676)

21:42:42.514 Error AndroidRuntime 2797   at android.app.LoadedApk$ReceiverDispatcher$Args.run(LoadedApk.java:755)

21:42:42.514 Error AndroidRuntime 2797   at android.os.Handler.handleCallback(Handler.java:615)

21:42:42.514 Error AndroidRuntime 2797   at android.os.Handler.dispatchMessage(Handler.java:92)

21:42:42.514 Error AndroidRuntime 2797   at android.os.Looper.loop(Looper.java:137)

21:42:42.514 Error AndroidRuntime 2797   at android.app.ActivityThread.main(ActivityThread.java:4745)

21:42:42.514 Error AndroidRuntime 2797   at java.lang.reflect.Method.invokeNative(Native Method)

21:42:42.514 Error AndroidRuntime 2797   at java.lang.reflect.Method.invoke(Method.java:511)

21:42:42.514 Error AndroidRuntime 2797   at com.android.internal.os.ZygoteInit$MethodAndArgsCaller.run(ZygoteInit.java:786)

21:42:42.514 Error AndroidRuntime 2797   at com.android.internal.os.ZygoteInit.main(ZygoteInit.java:553)

21:42:42.514 Error AndroidRuntime 2797   at dalvik.system.NativeStart.main(Native Method)

21:42:42.514 Warning ActivityManager 271    Force finishing activity com.google.android.youtube/.app.honeycomb.tablet.WatchActivity

21:42:42.619 Debug dalvikvm 358  GC_CONCURRENT freed 453K, 74% free 6589K/24775K, paused 18ms+3ms, total 61ms

21:42:43.010 Warning ActivityManager 271  Activity pause timeout for ActivityRecord{42941328 com.google.android.youtube/.app.honeycomb.tablet.WatchActivity}

21:42:52.510 Warning ActivityManager 271  Launch timeout has expired, giving up wake lock!

[Guest ejtagle]

Eduardo,

Just got this by playing video. Perhaps it says something to you:

21:42:27.347 Debug dalvikvm 2797  WAIT_FOR_CONCURRENT_GC blocked 51ms

21:42:28.073 Info OMXCodec 103  [OMX.google.h264.decoder] video dimensions are 480 x 368

21:42:28.103 Info OMXCodec 103  [OMX.google.h264.decoder] Crop rect is 480 x 360 @ (0, 0)

21:42:28.103 Info OMXCodec 103  [OMX.google.h264.decoder] video dimensions are 480 x 368

21:42:28.103 Info OMXCodec 103  [OMX.google.h264.decoder] Crop rect is 480 x 360 @ (0, 0)

21:42:28.103 Info SoftAAC2 103  Reconfiguring decoder: 44100 Hz, 2 channels

21:42:28.125 Debug audio_hw_primary 103  start_output_stream: device:0, rate:44100

21:42:29.635 Debug dalvikvm 271  GC_CONCURRENT freed 1069K, 24% free 10470K/13703K, paused 2ms+7ms, total 53ms

21:42:31.803 Debug dalvikvm 2797  GC_CONCURRENT freed 1954K, 11% free 19899K/22215K, paused 13ms+8ms, total 90ms

21:42:31.858 Debug dalvikvm 2797  WAIT_FOR_CONCURRENT_GC blocked 65ms

21:42:36.921 Debug dalvikvm 2797  GC_FOR_ALLOC freed 1365K, 12% free 19872K/22343K, paused 54ms, total 54ms

21:42:42.487 Debug dalvikvm 2797  --- called into dvmAbstractMethodStub

21:42:42.487 Debug AndroidRuntime 2797  Shutting down VM

21:42:42.514 Warning dalvikvm 2797  threadid=1: thread exiting with uncaught exception (group=0x41fa9300)

21:42:42.514 Error AndroidRuntime 2797  FATAL EXCEPTION: main

21:42:42.514 Error AndroidRuntime 2797  java.lang.AbstractMethodError: abstract method not implemented

21:42:42.514 Error AndroidRuntime 2797   at android.media.MediaRouter$Callback.onRouteVolumeChanged(MediaRouter.java)

21:42:42.514 Error AndroidRuntime 2797   at android.media.MediaRouter.dispatchRouteVolumeChanged(MediaRouter.java:592)

21:42:42.514 Error AndroidRuntime 2797   at android.media.MediaRouter.systemVolumeChanged(MediaRouter.java:603)

21:42:42.514 Error AndroidRuntime 2797   at android.media.MediaRouter$VolumeChangeReceiver.onReceive(MediaRouter.java:1676)

21:42:42.514 Error AndroidRuntime 2797   at android.app.LoadedApk$ReceiverDispatcher$Args.run(LoadedApk.java:755)

21:42:42.514 Error AndroidRuntime 2797   at android.os.Handler.handleCallback(Handler.java:615)

21:42:42.514 Error AndroidRuntime 2797   at android.os.Handler.dispatchMessage(Handler.java:92)

21:42:42.514 Error AndroidRuntime 2797   at android.os.Looper.loop(Looper.java:137)

21:42:42.514 Error AndroidRuntime 2797   at android.app.ActivityThread.main(ActivityThread.java:4745)

21:42:42.514 Error AndroidRuntime 2797   at java.lang.reflect.Method.invokeNative(Native Method)

21:42:42.514 Error AndroidRuntime 2797   at java.lang.reflect.Method.invoke(Method.java:511)

21:42:42.514 Error AndroidRuntime 2797   at com.android.internal.os.ZygoteInit$MethodAndArgsCaller.run(ZygoteInit.java:786)

21:42:42.514 Error AndroidRuntime 2797   at com.android.internal.os.ZygoteInit.main(ZygoteInit.java:553)

21:42:42.514 Error AndroidRuntime 2797   at dalvik.system.NativeStart.main(Native Method)

21:42:42.514 Warning ActivityManager 271	Force finishing activity com.google.android.youtube/.app.honeycomb.tablet.WatchActivity

21:42:42.619 Debug dalvikvm 358  GC_CONCURRENT freed 453K, 74% free 6589K/24775K, paused 18ms+3ms, total 61ms

21:42:43.010 Warning ActivityManager 271  Activity pause timeout for ActivityRecord{42941328 com.google.android.youtube/.app.honeycomb.tablet.WatchActivity}

21:42:52.510 Warning ActivityManager 271  Launch timeout has expired, giving up wake lock!

libmedia patch ? ... You are not using the nvidia hw accelerated decoder?

media-codec.xml?

audio_policy.conf?

media-profiles.xml ?

One of those things should be the cause of your problems.. ;)

[Guest Scanno]

libmedia patch ? ... You are not using the nvidia hw accelerated decoder?

media-codec.xml?

audio_policy.conf?

media-profiles.xml ?

One of those things should be the cause of your problems.. ;)

Sorry about that... thought i had included the libmedia patch already, but did not.... Should get more sleep :wacko:

[Guest ejtagle]

Sorry about that... thought i had included the libmedia patch already, but did not.... Should get more sleep :wacko:

No problems... It is very important to be able to duplicate what brucelee did. We want to make sure everything works :)

[Guest brucelee666]

Scanno (or anyone else),

Attached is my ver2.0 of the wifi jellybean patch, hopefully wifi connects and you get the bars showing for signal strength as I have - as before copy over the files and replace in relevant dirs.

You probably do this but you just need to "make -j1 wpa_cli wpa_supplicant libwpa_client" and copy these 3 files into appropriate system dirs and repackage and flash your image to hopefully get it working.

edit - re. youtube videos in aosp browser(not tried other browsers), what I have found after watching music videos for the last hour or so is that if you let the video play it will usually crash about half way through (white background grey arrow) the only thing logcat shows is the bufferqueue for the browser cancels the buffer as its not owned by the client.

Now if the video starts playing and you change the setting from 360 to 480 or 720 the video will usually play through fine with no issues (I have watched a number of videos all the way through by doing this), also if the video starts you can change to 480 then back to 360 in quick succession and the video will play normally.

It seems to be the first play without changing the setting, if you change the setting then everything plays fine (all the config profiles seem to match the settings correctly) don't have any reasons just reporting as you will prbably see it when you test it.

jellybean_patchv2.rar

Edited July 20, 2012 by brucelee666

[Guest ejtagle]

Well, let's try an audio fix again :)

Attached the rel15r7 kernel patch for shuttle. The only changes are to the /sound/soc folder. Specifically, compared to the previous patch, i have modified the function tegra_alc5624_jack_notifier inside tegra_alc5632.c . That is the function that sends the notification of headphone plugging and unplugging. There was a bug on the previous implementation, i hope i was able to fix it.

If you notice the output is switched (i mean, speaker output when headphones plugged in, or headphone output when no headphone plugged (i know, obviously no sound ;) ... it means that i have missed the headphone detect gpio polarity. If that is the case, please add to the tegra_alc5624_hp_jack_gpio structure, after the line

.debounce_time = 150,

the following line:

.invert = 1,

That line inverts the detection ..

You will also notice that under special cases android outputs sound to both speakers and headphones. That is NOT a bug. It is an Android feature ;) ... Ringtones are an example of that. There are ways to control that :) ... But those are user settings, not something we should worry about.

And you will notice android controls the sound output, as it should.

Hope we get it right this time.

PD: The latest codec driver allows us to control the codec hw equalizer. You can try that feature (i did not test it) , by using tinymix, look for the eq mode settting and try to change it ;)

rel-15r7.rar

[Guest Scanno]

Just recompiled the kernel with the /sound/soc changes. Seems not to make any difference. No sound throguh speaker.

Have to stop.. already late for work.

Perhaps i missed something. I have a look again in the evening

Edited July 20, 2012 by Scanno

[Guest foxsen]

Thank you for your information. But as I understand it, Richardmlea mainly provides a method to fix firmwares with wrong version 99.4.8.0, that will lead to completely unused touchscreen, and reboot/reflash normal roms won't solve it. My problem is a bit different: the firmware version is ok, usable most of the time, but fail sometimes; shutdown it, power on, the touch screen will be back.

Like I said in last post, at that time it seems the touch screen controller does not response to I2C bus requests.

The nvidia 15r7 Release note says that it is a known issue. In fact, not only the It7260 has 'no acknowledge' issue, I have also seen the power regulator (addr 4-0034) reports it. Most of time the device will continue to work even with this message, but there are bad times.

I have tried to build the driver as module, when the touchscreen becomes unresponsive, rmmod and insmod seems to revive the screen. The it8260.c need some patches to be safely insmod/rmmod.

Edited July 20, 2012 by foxsen

[Guest rvdgeer]

Sorry to interrupt this kernel dev thread...

On the dev thread for Jellybean for the HTC One X they experience the same kind of problems with audio: not being able to play it over a wirde headset...

However, a lot of user reported it does work when using a specific app named 'Wired Headphone/Headset Fix', available in the Google Play Store:

https://play.google.com/store/apps/details?id=com.woodslink.android.wiredheadphoneroutingfix

Might help a little on troubleshooting...

[Guest ejtagle]

Don't worry about audio. I am sure we will get it to work. I didn't have enough time to test it on the actual shuttle hw, as i am developing for a different device (also tegra2 based), and switching development environment to try the kernel for the shuttle takes time.

All i can say is that i was able to make speaker/headphone switch work on my device, so, i guess we won't have problems making it work in the p10an01.

We must differenciate between a kernel bug (codec (=alc5624.c) and/or glue logic (=tegra_alc5624.c), or an android bug (framework) or the audio driver in android (=audio_hw.c)

The audio_hw.c is nearly exactly the same as the one i am using, so, i guess there is no bug there. We are using the same framework, so, i also expect no bug there. The glue logic y nearly the same but not quite, so it could eb the problem... and the alc5624.c is completely different to the one i am using in my device (it is a P10an10) ...

Steps to diagnose the problem:

Very easy:

using logcat, make sure that when a headphone is plugged in, the last line says so:

audio_hw_primary: Headphone out:Y , Speaker out:N, HDMI out:N

And when you unplug it, it should say

audio_hw_primary: Headphone out:N , Speaker out:Y, HDMI out:N

Don't try this with ringtones. Ringtones, according to Android specs, will always be output to speakers (and also to headphones, if plugged in). Use the music application or youtube to try this.

That message reflects what the Android framework is requesting

If you find out that the headphone state is inverted, then you need to add the .invert line i posted earlier to the tegra_alc5624.c

With that working, the only remaining culprit should eb the audio codec driver itself. You could even try to use an older codec , but with the new tegra_alc5624.c ...

[Guest Scanno]

Never mind .. I should first check everythg and recompile to see if I have all the patches.

Edited July 20, 2012 by Scanno

[Guest grnunn]

Do we think that the issue is the codec being written to the wrong state? (As in the register C7 issue) Or do we think the codec driver is getting out of sync with the codec?

From the logcat from a while ago the codec driver looked as though it tried to write to all the chip registers at startup. If they haven't put a lockout on the test registers for the device then you could be putting the device in an "odd" state. Is there any way of limiting the registers that the codec driver writes at startup to defined registers rather than the whole address space? The reason I say this is because the driver was writing all zeros to the higher register addresses when (as is the case of C7) this was not correct.

Edited July 20, 2012 by grnunn

[Guest ejtagle]

Do we think that the issue is the codec being written to the wrong state? (As in the register C7 issue) Or do we think the codec driver is getting out of sync with the codec?

From the logcat from a while ago the codec driver looked as though it tried to write to all the chip registers at startup. If they haven't put a lockout on the test registers for the device then you could be putting the device in an "odd" state. Is there any way of limiting the registers that the codec driver writes at startup to defined registers rather than the whole address space? The reason I say this is because the driver was writing all zeros to the higher register addresses when (as is the case of C7) this was not correct.

You are right .. the 2.6.39 codec driver was writing only to specific registers... The newer one, due to different structures used by the alsa SoC audio layer, use a different structure layout that prevents that. I think there is a way to implement the previous behaviour, but with the newer structures. I will try to do it :)

[Guest Scanno]

I have rebuild JellyBean and checked all the patches. I have video working (HD video through Youtube app) and see in logcat that the nvidia display drivers are called. So i guess it is working. The latest patches from Brucelee666 are working and the wifi signal levels are shown. The wifi connection is not saved and are lost after a reboot so you have to enter the wpa key again.

[Guest Oldbarzo]

I have rebuild JellyBean and checked all the patches. I have video working (HD video through Youtube app) and see in logcat that the nvidia display drivers are called. So i guess it is working. The latest patches from Brucelee666 are working and the wifi signal levels are shown. The wifi connection is not saved and are lost after a reboot so you have to enter the wpa key again.

Hi scanno

Looking forward to your first Alpha for jellybean on Vega etc. Ready to test if required.

Stransky

[Guest Scanno]

You are right .. the 2.6.39 codec driver was writing only to specific registers... The newer one, due to different structures used by the alsa SoC audio layer, use a different structure layout that prevents that. I think there is a way to implement the previous behaviour, but with the newer structures. I will try to do it :)

Eduardo I am trying to port the power menu from cm10 so I downloaded the frameworks/base from them. During comparing I saw that they added some small stuff about tegra / alsa . Is that perhaps something to look into?

[Guest ejtagle]

Eduardo I am trying to port the power menu from cm10 so I downloaded the frameworks/base from them. During comparing I saw that they added some small stuff about tegra / alsa . Is that perhaps something to look into?

I think we should not add the alsa patches. I have seen that some roms are using older audio_hw.c that used alsa instead of tinyalsa. Google does not recommend using alsa anymore. The problems we have right now with audio are kernel related, not android related. I know because on the p10an10 i managed to make the audio fully operational (proper speaker and headphone support), and i am using nearly the same audio_hw.c android module i posted for the p10an01.

The main difference between x01 and x10 is the kernel, not the android framework (btw, if you ask me if i will release the x10 rom, probably i won't. Reason is that shuttle never released the kernel source code for that device. I managed to reverse engineer most of the drivers, but, something is missing, and i cannot figure it out. Bluetooth does not work, and wifi is unstable... I think it is something (probably a power supply) that i fail to enable, but, without schematics, or at least kernel sources, it is nearly impossible to figure it out)

All the other things are working. That is the device i am backporting things from to the x01 kernel.

And, i can tell x10 audio works beautifully. And, as we had 2.6.39 x01 audio working, i just think the problem lies in some of the porting to the 3.1 kernel.

As you told me that headphone sound is working, it must be something very silly. I lately tend to believe that grununn is right: I had to change the initialization sequence of the alc5624. Now, we touch all the codec registers, and previously, we touched just a few. Maybe we are touching a register that is disabling the speaker amplifier. I will undo that change to see if that is the problem :)

[Guest Scanno]

I think we should not add the alsa patches. I have seen that some roms are using older audio_hw.c that used alsa instead of tinyalsa. Google does not recommend using alsa anymore. The problems we have right now with audio are kernel related, not android related. I know because on the p10an10 i managed to make the audio fully operational (proper speaker and headphone support), and i am using nearly the same audio_hw.c android module i posted for the p10an01.

The main difference between x01 and x10 is the kernel, not the android framework (btw, if you ask me if i will release the x10 rom, probably i won't. Reason is that shuttle never released the kernel source code for that device. I managed to reverse engineer most of the drivers, but, something is missing, and i cannot figure it out. Bluetooth does not work, and wifi is unstable... I think it is something (probably a power supply) that i fail to enable, but, without schematics, or at least kernel sources, it is nearly impossible to figure it out)

All the other things are working. That is the device i am backporting things from to the x01 kernel.

And, i can tell x10 audio works beautifully. And, as we had 2.6.39 x01 audio working, i just think the problem lies in some of the porting to the 3.1 kernel.

As you told me that headphone sound is working, it must be something very silly. I lately tend to believe that grununn is right: I had to change the initialization sequence of the alc5624. Now, we touch all the codec registers, and previously, we touched just a few. Maybe we are touching a register that is disabling the speaker amplifier. I will undo that change to see if that is the problem :)

Eduardo,

The .39 drivers are working in 3.1, except for the mic. In logcat it complains about the mono/stereo input and when you record something, you hear static. I am using thee .39 drivers in vegacream rc2. The old drivers do not work in jellybean. But perhaps using the ICS audio_hw.c with the tinyalsa we use in ICS might also work in jb?

[Guest ejtagle]

Trying to use ICS audio_hw.c on JB is a waste of time. It is completely incompatible. I had to rewrite most of that code. tinyalsa has nothing to do with this problem.

The limitations you see regarding mic on 3.1 are not caused by the audio_hw,c or tinyalsa. It is caused by some of nvidia modifications to the kernel...

The exact problem is that nvidia removed from the i2s driver (this driver is part of the linux kernel) the mono output / input capability, Now, the only supported i2s mode is stereo.

I2s is the Tegra2 hardware port used to send or receive digital audio to/from the audio codec (alc5624.c) (the reponsible of converting digital audio to analog and amplify it to drive speakers/headphones).

But, the old tegra_alc5624.c we used in the 2.6.39 kernel expects i2s to support both mono and stereo. So, when someone (the Android framework) tries to capture monophonic audio from the mic, the tegra_alc5624.c file tries to configure the i2s tegra port for mono input. But, as the nvidia supplied driver does not support it anymore, it fails (and you will see the problem in dmesg).

The new tegra_alc5624.c implements a workaround for this problem, so when the framework tries to capture in mono using the tegra i2s port , instead of using the i2s mode of the port , we use another mode called dspa that supports mono capture. Our codec , the alc5624 also supports that mode, so it should work.

That is the reason i am trying to fix the alc5624.c / tegra_alc5624.c combination. Without that, we won't get proper audio support working :( ... The other possibility is to implement a workaround in the audio_hw.c we wrote for JB, transforming mono capture requests in stereo capture requests, and then using software to convert the stereo samples to mono. I just prefer to fix the kernel side, rather than implement such strange workarounds at the user level...

Attached a new attempt at fixing this. I reverted most of the changes to the alc5624.c regarding initial register initialization. Now it behaves much closer to the working alc5624.c from the 2.6.39 kernel... And we don't touch reserved codec registers anymore!

Also included an slight tweak to the touchscreen driver. If at some point it detects that the touchscreen driver has died, it resets it and reinits it. It can even power it down and up. Hopefully this will fix the remaining touchscreen "locks"...

Sorry for not telling about the exact audio capture problem cause... I simply had no time :(

edit1: Reuploaded the patch with a small improvement in the touchscreen driver. I was forgetting to reenable interrupts after the touchscreen reset when the driver tries to recover it when it detects it has died..

rel-15r7-v2.rar

Edited July 21, 2012 by ejtagle

[Guest Scanno]

Trying to use ICS audio_hw.c on JB is a waste of time. It is completely incompatible. I had to rewrite most of that code. tinyalsa has nothing to do with this problem.

The limitations you see regarding mic on 3.1 are not caused by the audio_hw,c or tinyalsa. It is caused by some of nvidia modifications to the kernel...

The exact problem is that nvidia removed from the i2s driver (this driver is part of the linux kernel) the mono output / input capability, Now, the only supported i2s mode is stereo.

I2s is the Tegra2 hardware port used to send or receive digital audio to/from the audio codec (alc5624.c) (the reponsible of converting digital audio to analog and amplify it to drive speakers/headphones).

But, the old tegra_alc5624.c we used in the 2.6.39 kernel expects i2s to support both mono and stereo. So, when someone (the Android framework) tries to capture monophonic audio from the mic, the tegra_alc5624.c file tries to configure the i2s tegra port for mono input. But, as the nvidia supplied driver does not support it anymore, it fails (and you will see the problem in dmesg).

The new tegra_alc5624.c implements a workaround for this problem, so when the framework tries to capture in mono using the tegra i2s port , instead of using the i2s mode of the port , we use another mode called dspa that supports mono capture. Our codec , the alc5624 also supports that mode, so it should work.

That is the reason i am trying to fix the alc5624.c / tegra_alc5624.c combination. Without that, we won't get proper audio support working :( ... The other possibility is to implement a workaround in the audio_hw.c we wrote for JB, transforming mono capture requests in stereo capture requests, and then using software to convert the stereo samples to mono. I just prefer to fix the kernel side, rather than implement such strange workarounds at the user level...

Attached a new attempt at fixing this. I reverted most of the changes to the alc5624.c regarding initial register initialization. Now it behaves much closer to the working alc5624.c from the 2.6.39 kernel... And we don't touch reserved codec registers anymore!

Also included an slight tweak to the touchscreen driver. If at some point it detects that the touchscreen driver has died, it resets it and reinits it. It can even power it down and up. Hopefully this will fix the remaining touchscreen "locks"...

Sorry for not telling about the exact audio capture problem cause... I simply had no time :(

edit1: Reuploaded the patch with a small improvement in the touchscreen driver. I was forgetting to reenable interrupts after the touchscreen reset when the driver tries to recover it when it detects it has died..

.

Eduardo,

Thanks for the explanation. Donot worry about not explaining. I am not sure if i would completely understand it. I have never programmed at low level. The most technical program i ever wrote was writing a communication programm using the IP stack and talking RFC821 / RFC822 (a.k.a SMTP)

Ok your latest kernel pacth does not work. No sound through speakers. But also mono sound through heasphones now.

Tried my previous kernel and i get stereo sound through headphones.

Also i have the powermenu ported and the reboot / reboot recovery is added to the framework and working.

Edited July 21, 2012 by Scanno

[Guest Scanno]

Eduardo,

I just got sound from speakers... that is good, but it is only mono, just like the headphones.

Only right channel is working. At first speakers are not working, but after inserting and removing the headphone jack, speakers are working.

Trying out Google Now, but the mic is producing static and cannot be used.

Looking good, we are getting there.

Edited July 21, 2012 by Scanno

[Guest brucelee666]

For info looks like google have pushed 4.1.1r1.1, r2 and r3 to git.

Have not had a chance to test the new kernel stuff will be later today before I get a chance, apart from audio I think the other main sticking point is the no access for /data by the settings app to either read config.txt or write wifi connection details.

Check logcat and you will see the 2 errors when it tries to do this the only thing it says is noaccess, when it tries to write it gives a delayed disk write - unable to write data error, any thoughts.

[Guest Scanno]

For info looks like google have pushed 4.1.1r1.1, r2 and r3 to git.

Have not had a chance to test the new kernel stuff will be later today before I get a chance, apart from audio I think the other main sticking point is the no access for /data by the settings app to either read config.txt or write wifi connection details.

Check logcat and you will see the 2 errors when it tries to do this the only thing it says is noaccess, when it tries to write it gives a delayed disk write - unable to write data error, any thoughts.

Repo syncing r3.

Regarding write errors, i have been fiddling with the init.rc / init.harmony.rc but so far no solution for that problem. Perhaps there is an error in the framework and r3 solves it. Anyway will see.