What is packing in structure and how compiler adds padding between member elements?

Structure members are placed one after another in the increasing order of memory. Now a CPU with 32bit data bus width is optimized to access 32bit or 4byte at a time and thus compiler arranges structure members in such a way that it maintains 32bit memory alignment. For 64bit this alignment is 64bit or 8bytes. Structure packing and padding extra bytes in structure members comes into picture.

Packing is an option in compiler to define how the members of a structure will be aligned in memory. It is defined via pre-processor #pragma pack(). For 32bit compiler pack(4) is default and similarly for 64bit pack(8) is default. Now we are instructing compiler to align members in 4byte memory spaces then what happen when we have some members with size less than 4. Here comes padding in picture. Some extra bytes are used by compiler to pad these gaps. Pack(1) is called tight packing. We do not have any type less than 1byte size thus this means no padding for pack(1).

#pragma pack(1)
struct pack_test_t {
  unsigned char m_byte;
};
sizeof(struct pack_test_t) is 1

#pragma pack(4)
struct pack_test_t {
  unsigned char m_byte;
};
sizeof(struct pack_test_t) is 4 [sizeof(m_byte) + sizeof(3 padding bytes)]

About our authors: Team EQA

You have viewed 1 page out of 248. Your C learning is 0.00% complete. Login to check your learning progress.

‹ # ›