Hi, I have few questions :
1.
I am running the OMX decode demo in SDK 5.1.1.80 omx_05_01_01_80/packages/ti/omx/demos/decode, and made some modification on it.
Originally, this demo runs with one decode component only, and it works fine on my EVM board.
Then I modified it to run with two decode components simultaneously, and it works fine with two streams of 1280x720 resolution.
However, when I try to decode one stream with 1280x720 and one stream with 1920x1080, it just stuck without any error message.
More specifically,
the first decoder : ....OMX_CommandPortEnable fot I/O ---> OMX_AllocateBuffer ---> Decode_WaitForState (OMX_StateIdle) correctly
the second one : ....OMX_CommandPortEnable fot I/O ---> OMX_AllocateBuffer ---> Decode_WaitForState (OMX_StateIdle) hung up at TIMM_OSAL_EventRetrieve
Is this caused by some out of memory issue ?
If yes, how can I extend the memory space or modify the memory map? or there is still something I hadn't taken care of ?
I have SDK 5.1.1.80 and SDK 5.0.0.11, and DM816x EVM Revision E.
2.
Is it possible to migrate the decode demo from SDK 5.1.1.80 to SDK 5.0.0.11 ?
Cause there are source codes of OMX components in SDK 5.0.0.11, and I might wish more control over this decode demo.
3.
I have seen a function IvaScheduler_asssignStaticResource in SDK 5.0.0.11 omx_05_00_00_11\packages\ti\omx\comp\scheduler\src\ivascheduler.c
Does this mean that I could assign as many decode components as possible, and need not care about which decoder uses which IVAHD ?
Just want to clarify it.
I really appreciate for any information regarding to my questions.
Thanks,
Yuanchi
