TDA4VM: How to efficiently exchange large data between A72 and MCU3_0

liu shen

Part Number: TDA4VM

Hi,

As described by “8.Developer Notes” -> "8.4. Developing IPC applications" -> "8.4.6. Performance and Latency":

To reduce latencies and / or to send larger buffer it is recommended to pass a pointer/handle/offset to a larger shared memory from ION heapCode

However, I did not find any manual about ION heap. How do I use it between A72 and R5f(MCU3 0)?

Thanks!

over 4 years ago

0 Stanley Liu over 4 years ago

TI__Guru 50045 points

Hi,

Please refer to the below FAQ first. It provides a simple app to follow.

https://e2e.ti.com/support/processors-group/processors/f/processors-forum/909695/faq-tda4vm-makefile-based-project-which-shows-a72-c7x-r5f-setup-along-with-ipc-and-buffer-passing-between-linux-rtos

Regards,
Stanley

0 liu shen over 4 years ago in reply to Stanley Liu

Prodigy 30 points

Thanks for your reply.

This seems to be a ipc demo，there is limitation with message payload size( max 512 bytes).

But I want to transfer more than 2K bytes at a time.

0 Suman Anna over 4 years ago in reply to liu shen

TI__Guru** 111215 points

Hi Liu Shen,

The above example does use the VisionApps API to allocate a buffer (example uses 32 bytes) from a DDR dma-heap carveout, but you can allocate larger memory.

There is no message payload size restriction when passing buffer pointers. The 512 bytes is the maximum rpmsg transport buffer size with a maximum payload of 496 bytes.

regards

Suman

Processors

Processors forum

TDA4VM: How to efficiently exchange large data between A72 and MCU3_0