TI OpenMAX Integration Layer project

This is a TI implementation of the  of the OpenMAX Integration Layer
(IL) API ver. 1.1, specified by the Khronos group (see http://www.khronos.org/openmax).

This Unified OMX is composed of:
- audio component(s)
  - OMX audio decoder
	- MP3 extension role
	- AAC extension role
- system component(s)
  - OMX Core
  - LCML (Linux Common Multimedia Library)
  - OMX Base (OMX base component)
  - OSAL (Operating system abstraction layer API)

The first version of UOMX was developed by TI India MM platforms team
Authors:
Aditya Monga
Radha Purnima Dasari

Introduction

The intent was to prototype the generic OMX-BASE implementation on the
OMAP3 platform and then extend the same to audio components on the OMAP4
platform running the Linux OS.

Why OMX-BASE and the associated changes?

Two generic components for audio encode and audio decode have been derived
from the generic "base" implementation of  OMX1.1.

The OMX-BASE was initially developed for eCosto on the Nucleus RTOS, but is
written in such a way that it can be easily ported onto other operating
systems with a rewrite of the OSAL compoment.

With the OMX-BASE writing a new component is less tedious since the
component writer has to only think about implementing the codec specific
aspects of the OMX component.

Location of package

You can download the package from:
http://dev.omapzoom.org/?p=abraham/audio-omx.git;a=summary

Configuration Settings

-  Linux® OS: Linux 24.0.9 2.6.31-rc1
-  Tools: arm-none-linux-gnueabi-gcc (Sourcery G++ Lite 2008q3-72) 4.3.2
-  GNU Make version 3.81
-  File System: Virtio RAMDISK
-  Hardware   : Virtio 1.09a patch 9

Highlights

- This is the first release of UOMX audio components for OMAP4.
- MP3/AAC decoding in file to file mode.
- The test application "tiomxplayer" is provided to test AUDIO.DECODE component
  with MP3/AAC role.

Known Issues/Limitations

- Currently MP3 and AAC decoding is supported only in stream mode.
- MP3 UOMX test cases has only been validated using a static baseimage provided
   by the SN team.
- AAC UOMX test cases has only been validated using a static baseimage provided
   by the SN team.

Build Instructions

The code has NOT been integrated into clear case.
You'll also need ALSA utils and libraries in order to render the test
application through alsa (alsa libs are also provided).
For this release L23.17 (LINUX-MMAUDIO_RLS_3.23P5) clearcase baseline was used.

As for this relase UOMX binaries are provided and ready to be integrated into
a ramdisk. You'll find them on binaries folder.

Genral instruction to compile UOMX on clearcase....

1.	Clone the UOMX git project to your local machine:
	$ git clone git://dev.omapzoom.org/pub/scm/abraham/audio-omx.git
2. 	In clearcase export these variables:
	$ export OMXROOT=/vobs/wtbu/OMAPSW_MPU/linux
	then create a PKGDIR direcotry and export it's path
	$ export PKGDIR=/home/<user>/file_system
	$ export CROSS=arm-none-linux-gnueabi-
	$ export OMX_RESOURCEMANAGER_DISABLE=1
	$ export PATH=$PATH:<path_to_your>/arm-2008q3-72/bin/

Building OMX-core
1.      Replace your /vobs/wtbu/OMAPSW_MPU/linux/system/src/openmax_il/omx_core
        files with the one you just clone from the git tree audio-omx/system/src/openmax_il/omx_core/
2. 	Go to /vobs/wtbu/OMAPSW_MPU/linux/system/src/openmax_il/omx_core/ and type:
	$ make clean all

Building the LCML
Due to LCML dependencies on Bridge you'll need to copy the Brige libraries needed to compile LCML
1. 	Create this directory structure:
	$ mkdir $PKGDIR/target/include/dspbridge/
2. 	Copy /vobs/wtbu/OMAPSW_MPU/dspbridge/mpu_api/inc/* to your $PKGDIR/target/include/dspbridge/
	NOTE: You'll also need to have the bridge shared libraries (libbridge.so and libbridge.so.2)
		  in $PKGDIR/target/include/dspbridge/lib refer to Bridge Release notes on how to complile
		  these libraries.
3. 	Replace your /vobs/wtbu/OMAPSW_MPU/linux/system/src/openmax_il/lcml/ files  with the one you just
	clone from the git tree audio-omx/system/src/openmax_il/lcml/
4. 	Go to /vobs/wtbu/OMAPSW_MPU/linux/system/src/openmax_il/lcml/ and type:
	$ make clean all

Building the OSAL layer
1. 	Copy to your /vobs/wtbu/OMAPSW_MPU/linux/system/src/openmax_il/ directory the mm_osal directory you just
	clone from the git tree audio-omx/system/src/openmax_il/mm_osal
2. 	Go to /vobs/wtbu/OMAPSW_MPU/linux/system/src/openmax_il/mm_osal/ and type:
	$ make clean all

Building the OMX-Base
1. 	Copy to your /vobs/wtbu/OMAPSW_MPU/linux/system/src/openmax_il/ directory the omx_base directory you just
	clone from the git tree audio-omx/system/src/openmax_il/omx_base
2. 	Go to /vobs/wtbu/OMAPSW_MPU/linux/system/src/openmax_il/omx_base/ and type:
	$ make clean all

Building the audio_decode OMX component
1. 	Copy to your /vobs/wtbu/OMAPSW_MPU/linux/audio/src/openmax_il directory the audio_decode
	 directory you just clone from the git tree audio-omx/audio_/src/openmax_il/audio_decode
2. 	Copy to your /vobs/wtbu/OMAPSW_MPU/linux/audio/src/openmax_il directory the mp3_dec
	 directory you just clone from the git tree audio-omx/audio_/src/openmax_il/mp3_dec
3. 	Copy to your /vobs/wtbu/OMAPSW_MPU/linux/audio/src/openmax_il directory the aac_dec
	 directory you just clone from the git tree audio-omx/audio_/src/openmax_il/aac_dec
4. 	Go to /vobs/wtbu/OMAPSW_MPU/linux/audio/src/openmax_il/audio_decode and type:
	$ make clean all

Running on Virtio

- After all compilation takes place copy the next libraries into your Virtio ramdisk
 a. Copy
	-OMAPS_MPU/linux/audio/src/openmax_il/audio_decode/src/libOMX.TI.AUDIO.DECODE.so,
	-OMAPS_MPU/linux/system/src/openmax_il/lcml/src/libLCML.so,
	-OMAPS_MPU/linux/system/src/openmax_il/omx_base/src/libOMX_Base.so,
	-OMAPS_MPU/linux/system/src/openmax_il/omx_core/src/libOMX_Core.so
	-OMAPS_MPU/linux/system/src/openmax_il/mm_osal/src/libmmosal.a
    into your ramdisk /usr/lib directory.

 b. Copy OMAPS_MPU/linux/audio/src/openmax_il/audio_decode/tests/tiomxplayer into your
	ramdisk /usb/bin directory

 c. Don't forget ot copy some MP3 test vectors as well.

 d. Boot into Virtio and load MP3 static baseimage.

 e. Run
    # tiomxplayer -o out.pcm <path_to>/<your_mp3_file>.mp3
	and wait for the test to finish, you'll see it creates the out.pcm file after.

 f. Play the out.pcm with ALSA
    # aplay out.pcm -f S16_LE -c <Number_of_Channels> -r <Sample_rate>