Hi edenli,
There are at least 2 methods i could think of right now:
- #1: quick in and dirty is to declare a reserved buffer in Simulink with a size you control. Since in Simulink you control the read/write into such buffer you could easily implement what you want with writes and reads to any specific index in the buffer.
Am saying is a quick and dirty since you do not control the start address of the buffer - depending on your Simulink modification that might change and you might need to re-adapt the start address of the DataMemWrite/Read Blocks (the offsets remains unchanged). If you use instead DataMemWrite/Read the standard Simulink Read/WriteStore blocks - then everything would be ok and you could forget about address changing due to adding new stuff in the model.
How to declare a 100 uint32 size buffer called myReservedSpace in Simulink:
How to implement the model
How to check the memory allocation:
- #2: the "profi" method would be to use a reserved memory area in the SRAM that have fixed start address and size.
For that you need to modify the Toolbox standard linker command files ( from here: {toolbox install dir}\rappid564xl\src\cw_specific_files\) by spliting the existing SRAM memory into 3 (right now there are 2 areas). So you need to add a new one: before int_sram or after int_sram. The stack_ram would be better to be kept last in order to avoid stack overflow issues.
Here is an example: cortex m3 - How can I make an empty section with GNU ld? - Stack Overflow
Once you have defined the area, you could take the start address of it an then using the DataMemWrite/Read you could control what you place in the that area.
Hope this helps!
Daniel
