Structure Size

A large number of processors require that integers larger than one byte must be aligned corresponding to their size, i.e. a 16-bit integer mut be aligned on an even address, and a 32-bit integer must be aligned in an address evenly dividable by 4.

If you have a struct containing one 32-bit integer and one 16-bit integer, the used size may be 4+2 = 6 bytes. However, to be able to create arrays of such a struct, the struct must be padded to 8 bytes or the 32-bit integer in the following struct would not be correctly aligned, i.e.:

struct {
    u32 a;
    u16 b;
} my_array[4];

will result in the following:

offset value
0      my_array[0].a
4      my_array[0].b
6      2-byte padding
8      my_array[1].a
12     my_array[1].b
14     2-byte padding
16     my_array[2].a
...

In your case, you have a struct that contains at least one 32-bit integer. Therefore, the struct will have a size that is 4*n byte large.

A large number of processors require that integers larger than one byte must be aligned corresponding to their size, i.e. a 16-bit integer mut be aligned on an even address, and a 32-bit integer must be aligned in an address evenly dividable by 4.

If you have a struct containing one 32-bit integer and one 16-bit integer, the used size may be 4+2 = 6 bytes. However, to be able to create arrays of such a struct, the struct must be padded to 8 bytes or the 32-bit integer in the following struct would not be correctly aligned, i.e.:

struct {
    u32 a;
    u16 b;
} my_array[4];

will result in the following:

offset value
0      my_array[0].a
4      my_array[0].b
6      2-byte padding
8      my_array[1].a
12     my_array[1].b
14     2-byte padding
16     my_array[2].a
...

In your case, you have a struct that contains at least one 32-bit integer. Therefore, the struct will have a size that is 4*n byte large.

To overcome the problem use the __packed attribute. More info: http://www.keil.com/support/man/docs/armcc/armcc_cjajddhb.htm