Assume that we have an array called:
uint8_t data_8_bit[2] = {color >> 8, color & 0xff};
The data is 16-bit color data. Our goal is to create an array called:
uint8_t data_16_bit[2*n];
Where n is actually the length of 16-bit data array. But the array data_16_bit cannot hold 16-bit values so therefore I have added a 2*n as array size.
Sure, I know that I can fill up the array data_16_bit by using a for-loop:
for(int i = 0; i < n; i++)
for(int j = 0; j < 2; j++)
data_16_bit[j*i] = data_8_bit[j];
But there must be a faster way than this?
memset or memcpy?
IMO the easiest one to optimize by the compiler (and very safe as well) is
void foo(uint16_t color, uint8_t *arr16, size_t n)
{
uint8_t data_8_bit[2] = {color >> 8, color & 0xff};
while(n--)
{
memcpy(arr16 + n * 2, data_8_bit, 2);
}
}