[DEV] ICS AOSP (from CM's source) Development Passion (N1) [Alpha8]

yumcax · Nov 19, 2011

Flashable .zip?

This is great, how hard would it be to create a flashable .zip file to install from?

drive2droad · Nov 19, 2011

I am installing in 3 .. 2 ... 1 ....

FrAsErTaG · Nov 19, 2011

yumcax said:
This is great, how hard would it be to create a flashable .zip file to install from?

not very, but why should the OP bother while he is just releasing alpha work?
fastboot is easy to use...

pconwell · Nov 19, 2011

I know you're busy, but...

Can you clean up your first post instead of just mangling edits all over the place. I'm trying to keep up with what you are doing, but it's just a hodge podge of edits. Something like:

Code:

Working:
Wifi
MMS
Bluetooth
TouchScreen

Code:

Not working:
3g (severe connect/disconnect issues)
Audio
Home button
Trackball
Camera
GPS

Near the top would be awesome.

Thanks.

skitty · Nov 19, 2011

I'm pretty sure drew Garen had the hard home button working in his sdk build. Maybe speak to him or have a sift through his build?
3g also is supposed to work but I never tested it as I could be bothered to enter my apn details

Sent from my Nexus One using xda premium

lim_ryan · Nov 19, 2011

any irc dev chat? or twitter follow dev? i unable to enter freenote

PeRKoniX · Nov 19, 2011

Cool that someone is developing this

though it's still kinda slow (I hope it will get better?) it would be nice to see that bottom bar gone since N1 already has softbuttons. I know that the android emulator has similar behavior (no on-screen buttons) since the emulator has it's own menu/home/back button. Maybe it is something to look into in the future of this project

Nice work! Thanks allot!

babijoee · Nov 19, 2011

pconwell said:
Can you clean up your first post instead of just mangling edits all over the place. I'm trying to keep up with what you are doing, but it's just a hodge podge of edits. Something like:

Code:

Working: Wifi MMS Bluetooth TouchScreen

Code:

Not working: 3g (severe connect/disconnect issues) Audio Home button Trackball Camera GPS

Near the top would be awesome.

Thanks.

I have to disagree. having the steps aswell as code encourages others to follow and help others who want to build/fix aosp. great work do far Texas.

DaXmax · Nov 19, 2011

Great work texasice! Im waiting for a full ver from u if u successfully port... But libaudio is the problem itself or??

Sent from my Nexus One using Tapatalk

texasice · Nov 19, 2011

DaxElectro said:
Great work texasice! Im waiting for a full ver from u if u successfully port... But libaudio is the problem itself or??

Sent from my Nexus One using Tapatalk

DJ_Jedi and I have been investigating. We think the driver is fine but something to do with how audioflinger is handling it.

Code:

Get this when I open the phone app and press the 2 button (as a test)

createTrack_l() Bad parameter: sampleRate 44100 format 1, channelMask 0x00000001 "for output 0x103b0 with format 1
AudioFlinger could not create track, status: -22
Error creating AudioTrack
createTrack_l() Bad parameter: sampleRate 44100 format 1, channelMask 0x00000001 "for output 0x103b0 with format 1
AudioFlinger could not create track, status: -22
AudioTrack->initCheck failed
ToneGenerator init failed 
Exception caught while creating local tone generator: java.lang.RuntimeException: Init failed
playTone: mToneGenerator == null, tone: 2

Changed the audioflinger source for more detail

Code:

[ 11-19 06:20:47.159    78:0x242 E/AudioFlinger ]
createTrack_l() Bad parameter: sampleRate 44100 m 44100 format 1 m 1, channelMask 0x00000001 " m 0x00000003 for output 0x103a8 with format 1

So it is either channelMask or mChannelMask
And we're stuck here right now.

davelis828 · Nov 19, 2011

texasice said:
Yea here's one that I wrote

Heres my logcat error if anyone knows what it means:

Code:

I/SystemServer( 152): Audio Service E/AudioFlinger( 78): createTrack_l() Bad parameter: sampleRate 44100 format 1, channelMask 0x00000001 "for output 0x103b0 with format 1 E/AudioTrack( 152): AudioFlinger could not create track, status: -22 E/SoundPool( 152): Error creating AudioTrack E/AudioFlinger( 78): createTrack_l() Bad parameter: sampleRate 44100 format 1, channelMask 0x00000001 "for output 0x103b0 with format 1 E/AudioTrack( 152): AudioFlinger could not create track, status: -22

repeat

Another logcat

I did adb logcat | grep Network

http://logcat.miui.us/506018

Here's one with logcat -b radio

http://logcat.miui.us/500034

thanx fella!

DaXmax · Nov 19, 2011

texasice said:

DJ_Jedi and I have been investigating. We think the driver is fine but something to do with how audioflinger is handling it.

Code:

Get this when I open the phone app and press the 2 button (as a test)

createTrack_l() Bad parameter: sampleRate 44100 format 1, channelMask 0x00000001 "for output 0x103b0 with format 1
AudioFlinger could not create track, status: -22
Error creating AudioTrack
createTrack_l() Bad parameter: sampleRate 44100 format 1, channelMask 0x00000001 "for output 0x103b0 with format 1
AudioFlinger could not create track, status: -22
AudioTrack->initCheck failed
ToneGenerator init failed 
Exception caught while creating local tone generator: java.lang.RuntimeException: Init failed
playTone: mToneGenerator == null, tone: 2

Changed the audioflinger source for more detail

Code:

[ 11-19 06:20:47.159    78:0x242 E/AudioFlinger ]
createTrack_l() Bad parameter: sampleRate 44100 m 44100 format 1 m 1, channelMask 0x00000001 " m 0x00000003 for output 0x103a8 with format 1

So it is either channelMask or mChannelMask
And we're stuck here right now.

This logcat is somehow similar to my fellas at Optimus One while porting ICS.... One of the users manage to get the audio working but....

Sent from my Nexus One using Tapatalk

Chaosz-X · Nov 19, 2011

Yes, we Desire-people get that error too

We tried to insert these files: http://recluster.net/ics4bravo/alllibs.zip

But it didn't work, but it seems we're having the exact same issue. Do any of you people here have an idea how to fix it? ;p

HairyMerkin · Nov 20, 2011

I wish someone would post an update.zip. I have no access to a linux box to compile!

babijoee · Nov 20, 2011

Okay i'm going to install 10.04..
11.10 sucks bad and re-download the source.

twistedddx · Nov 20, 2011

In one version of code I just skim read on CodeAurora, maybe not 4.0 code?:

channelCount is suppose to be an int.
If you do createTrack with channelCount set as 0 it will just set it to 2.
1 will be mono, 2 for stereo, all other values are invalid and will throw BAD_VALUE aka -EINVAL aka -22.

In theory 0x00000001 is valid(hopefully it's not being converted to a text string or something?!) but I saw a 0x00000003 also in your logs which would be invalid.

It can throw BAD_VALUE for more things than just setting a channelCount above 2.

I would start by proving you are REALLY setting channelCount as either 0, 1 or 2 and NOTHING else.

Edit: From other code i Just read on gitorious, is this 4.0 code?:
The only place I can find the same error output is in:
http://www.gitorious.org/0xdroid/fr...4a39b09278/libs/audioflinger/AudioFlinger.cpp

if (sampleRate != mSampleRate || format != mFormat || channelCount != mChannelCount) {
LOGE("createTrack_l() Bad parameter: sampleRate %d format %d, channelCount %d for output %p",
sampleRate, format, channelCount, mOutput);
lStatus = BAD_VALUE;
goto Exit;
}

I assume:
createTrack_l() Bad parameter: sampleRate 44100 m 44100 format 1 m 1, channelMask 0x00000001 " m 0x00000003 for output 0x103a8 with format 1

Is where you also printed the mChannelCount which was 0x00000003, but channelCount was at 0x00000001 and because they are not equal it throws the error.
Sounds like an AudioTrack object was created with mchannelCount as 3 instead of 1, then moments later you have channelCount as 1 and the code fails.

Why was the AudioTrack created with mchannelCount 3? I think if its created with no params it creates it as CHANNEL_OUT_STEREO
CHANNEL_OUT_STEREO = (CHANNEL_OUT_FRONT_LEFT | CHANNEL_OUT_FRONT_RIGHT),
CHANNEL_OUT_FRONT_LEFT = 0x4,
CHANNEL_OUT_FRONT_RIGHT = 0x8,

Which leads to 0x00001100 which is kind of close to, but entirely not, 0x00000011 which in hex would read 0x000003

So well I have no great ideas where the problem stems from. Track down when the AudioTrack gets created, if you are setting it up with or without params. If without it'll set as stereo unless you later go back to re-set the channelCount correctly to your desired 1.

JieeHD · Nov 20, 2011

Code:

AudioStreamOut* AudioHardware::openOutputStream(
        uint32_t devices, int *format, uint32_t *channels, uint32_t *sampleRate, status_t *status)
{
    { // scope for the lock
        android::Mutex::Autolock lock(mLock);

        // only one output stream allowed
        if (mOutput) {
            if (status) {
                *status = INVALID_OPERATION;
            } else { 
				LOGD("Output stream already exists, using existing channels and format"); 
				*status = NO_ERROR; 
			} 
		} 
		return mOutput; 
	}

        // create new output stream
        AudioStreamOutMSM72xx* out = new AudioStreamOutMSM72xx();
        status_t lStatus = out->set(this, devices, format, channels, sampleRate);
        if (status) {
            *status = lStatus;
        }
        if (lStatus == NO_ERROR) {
            mOutput = out;
        } else {
            delete out;
        }
    }
    return mOutput;
}

Hoping this should fix audio issues, will try it when I have time.

crazybyte · Nov 20, 2011

JieeHD said:

Code:

AudioStreamOut* AudioHardware::openOutputStream(
        uint32_t devices, int *format, uint32_t *channels, uint32_t *sampleRate, status_t *status)
{
    { // scope for the lock
        android::Mutex::Autolock lock(mLock);

        // only one output stream allowed
        if (mOutput) {
            if (status) {
                *status = INVALID_OPERATION;
            } else { 
				LOGD("Output stream already exists, using existing channels and format"); 
				*status = NO_ERROR; 
			} 
		} 
		return mOutput; 
	}

        // create new output stream
        AudioStreamOutMSM72xx* out = new AudioStreamOutMSM72xx();
        status_t lStatus = out->set(this, devices, format, channels, sampleRate);
        if (status) {
            *status = lStatus;
        }
        if (lStatus == NO_ERROR) {
            mOutput = out;
        } else {
            delete out;
        }
    }
    return mOutput;
}

Hoping this should fix audio issues, will try it when I have time.

In this code you're not relasing the mutex... i think that we have to relase it before return...

texasice · Nov 20, 2011

crazybyte said:
In this code you're not relasing the mutex... i think that we have to relase it before return...

Can you elaborate. With JieeHD code sound still does not work

crazybyte · Nov 20, 2011

I cant write any line of code because i doesn't know where to find info about the NDK api, in android. But if in some pice of code we do a lock(mutex&) it's mandatory to do a relase(mutex&). Otherwise it will make a deadlock...

In other topic we saw that someone found the 3.0 android source. It's good idea to check the code associated to the non working parts and see the difference with the CM7 kernel.

[DEV] ICS AOSP (from CM's source) Development Passion (N1) [Alpha8]

Senior Member

Senior Member

Senior Member

Senior Member

Senior Member

Senior Member

Senior Member

Retired Forum Moderator

Senior Member

Retired Recognized Developer

Senior Member

Senior Member

Senior Member

Senior Member

Retired Forum Moderator

Senior Member

Retired Recognized Developer

Senior Member

Retired Recognized Developer

Senior Member

Similar threads

Top Liked Posts