Rpmsg not running stable on m4

Yes this is a good patch overall and makes most sense, and should be the default code.

But the RO could be handy, perhaps? (Actual no physical ROM doesn’t matter; just the principle)

What if you can lock entire text segment with init data from say u-boot? Unlike locking say part of the data segment may be less appealing.

You would agree that the saving, code-wise, from removing the loading code rapidly tends to 0.

@dry it was actually my first time I had to go so deep in debugging and you actually helped me quite a bit. Thanks a lot, without you I wouldn’t have found it!

@stefan.tx No worries, I can understand that. Yes, by dropping the AT it should be fine for the remoteproc too. Actually a very elegant way to solve this. Nice patch!

Hello people, i have a working code for the rpmsg bare metal GitHub - warrierabhinav/rpmsglite_nanopb_imx8: This is a unique communication stack which uses RPMSG lite for heterogeneous multicore processing on top of which the data is serialized using NanoPb(smaller footprint of Google Protocol Buffers ) . This was developed and tested for the imx8 by Nxp . The board was Nitrogen8m by Boundary devices . Enjoy!