modlfo
commented
3 years ago There are two ways that you can do it depending on your needs.
In situations when I have to allocate or reserve a big chunk of memory, for example when creating a delay, I usually access it with external functions. For example:
// Test.vult
// Declare two external function to read and write the data
external writeToBuffer(index:int, value:real) "writeToBuffer";
external readFromBuffer(index:int) : value "readFromBuffer";// main.cpp
float* buffer;
#define DELAY_SIZE 1024
void initializeBuffer() {
buffer = (float*)malloc(sizeof(float) * DELAY_SIZE); // use your custom allocator
}
extern "C" {
void writeToBuffer(int index, float value) {
buffer[index] = value;
}
float readFromBuffer(int index){
return buffer[index];
}
}That way your Vult code can access the buffer.
The second option would be if you want to put all the memory used by your Vult code in the external memory. For example:
// Test.vult
// A function in Vult that requires memory
fun process(x:real) {
mem count = count + 1;
return count;
}// main.cpp
#include "test.h"
int main(void) {
// Allocate the memory used by the Vult code using your custom allocator
Test_process_type* processor = (Test_process_type*) malloc(sizeof(Test_process_type));
// then pass it to the Vult functions as follows.
Test__ctx_type_2_init(*processor);
return 0;
}
DatanoiseTV
I'm running Vult generated code on my ESP32, but internal RAM is limited to 520K (minus WiFi Stack etc), so in order to use more, I need to use dynamic memory allocation to use the external PSRAM.
This is how I do it in non-vult DSP code.
ps_malloc allocates memory on external PSRAM.
Do I need to write a custom generator?