Byte order functions on stm32 microcontroller - endianness

Question

Are there any written functions for htonll,ntohll,ntohl,htonl like this one i've found here on forum:
uint32_t ntohl(uint32_t const net) {
uint8_t data[4] = {};
memcpy(&data, &net, sizeof(data));
return ((uint32_t) data[3] << 0)
| ((uint32_t) data[2] << 8)
| ((uint32_t) data[1] << 16)
| ((uint32_t) data[0] << 24);
}
or could you just tell me data shifting scheme for each one and would write it just like this one above?
EDIT:
Are those correct:
uint32_t htonl(uint32_t const net) {
uint8_t data[4] = {};
memcpy(&data, &net, sizeof(data));
return ((uint32_t) data[0] << 0)
| ((uint32_t) data[1] << 8)
| ((uint32_t) data[2] << 16)
| ((uint32_t) data[3] << 24);
}
uint64_t htonll(uint64_t const net) {
uint8_t data[4] = {};
memcpy(&data, &net, sizeof(data));
return ((uint64_t) data[0] << 0)
| ((uint64_t) data[1] << 8)
| ((uint64_t) data[2] << 16)
| ((uint64_t) data[3] << 24);
}
uint64_t ntohll(uint64_t const net) {
uint8_t data[4] = {};
memcpy(&data, &net, sizeof(data));
return ((uint64_t) data[3] << 0)
| ((uint64_t) data[2] << 8)
| ((uint64_t) data[1] << 16)
| ((uint64_t) data[0] << 24);
}

Answer 1

This isn't STM32 specific but can be answered in general for all ARM MCU. Assuming you are using gcc, there are built-in functions like __builtin_bswap32.
So you can implement it as:
uint32_t htonl(uint32_t net)
{
return __builtin_bswap32(net);
}
uint16_t htons(uint16_t net)
{
return __builtin_bswap16(net);
}
uint64_t htonll(uint64_t net)
{
return __builtin_bswap64(net);
}
The resulting assembler code is very efficient:
htonl:
rev r0, r0
bx lr
htons:
rev16 r0, r0
uxth r0, r0
bx lr
htonll:
rev r3, r0
rev r0, r1
mov r1, r3
bx lr
If you declare the functions as inline, you can even save the function call.
BTW: In your code, the 64 bit versions are likely incorrect as they only return a 32 bit quantity.