I don't think this would be the cause of your problem but I would have thought thatshould bebased on what I see from other parts of the code. It seems like you are allocating ARRAY_BYTE_COUNT count of 32-bit pointers on the stack and not ARRAY_BYTE_COUNT bytes as expected.
Your synchronization problem looks like the first call to multicore_fifo_pop_blocking() works as expected but maybe the second one returns while the code on Core1 is still running. Could your try using a global g_core1Done variable to track the completion state of Core1 instead?
Code:
uint8_t* array[ARRAY_BYTE_COUNT] = {};Code:
uint8_t array[ARRAY_BYTE_COUNT] = {};Your synchronization problem looks like the first call to multicore_fifo_pop_blocking() works as expected but maybe the second one returns while the code on Core1 is still running. Could your try using a global g_core1Done variable to track the completion state of Core1 instead?
Code:
static volatile int g_core1Done = 0;...void multi_core_worker(void) { write_coords_row_wise(g_array, ARRAY_WIDTH, ARRAY_HEIGHT, ARRAY_HALF_WIDTH, ARRAY_WIDTH); g_core1Done = 1;}...void test_multi_core() { // Allocate a zero-initialized array on the stack. uint8_t array[ARRAY_BYTE_COUNT] = {}; multicore_reset_core1(); // Fill in the second half of the array on the second core. g_array = array; g_core1Done = 0; multicore_launch_core1(multi_core_worker); // Fill in the first half of the array on the first core. write_coords_row_wise(array, ARRAY_WIDTH, ARRAY_HEIGHT, 0, ARRAY_HALF_WIDTH); while (g_core1Done == 0) { // Busy wait for Core1 to complete. } ...}Statistics: Posted by adam_green — Mon Jan 01, 2024 2:48 am