How to store 8 bit character in 32 bit Data Flash memory?

Question

Is it possible to store 8 bit character to 32 bit Internal Data Flash in the Microcontroller by just using only 8 bit of space for each bytes?,by using a combination of 4 bytes which can be also recovered as individual bytes later. I'm using Nuc240 (ARM-M0) Microcontoller with 32 bit Data Flash and 512 bytes erase limit,read write limit is 32 bit.
Please explain a code template if possible.

Answer 1

Yes, by either array or structure.

char memory [4] = { 'a', 'b', 'c', 'd' };  // Uses 4-bytes (32-bits)

// memory [0] = 'a';
// memory [1] = 'b';
// memory [2] = 'c';
// memory [3] = 'd';

// ... store 'memory' in internal flash memory ...

Or

typedef struct
{
    char byte_0;
    char byte_1;
    char byte_2;
    char byte_3;
} memory_t;

memory_t memory; // Uses 4-bytes (32-bits );

memory.byte_0 = 'a';
memory.byte_1 = 'b';
memory.byte_2 = 'c';
memory.byte_3 = 'd';

// ... store 'memory' in internal flash memory ...

Answer 2

The standard way of storing data in flash at link-time when C programming, is to declare the variable as const.

const uint8_t array[4] = {0, 1, 2, 3};

Most compilers will then know to place this array in flash instead of RAM, but only if the variable also has static storage duration. Meaning that the above array must either be declared at file scope, or as static. See this for details.

Compilers also have non-standard syntax to place an array in a specific part of flash, which is what you have to use if the above won't work on your specific compiler. It can also be useful when for example dealing with eeprom/data flash.