Thread: question about bit fields?

when i watch a video about bit fileds to understand it
the teacher who made the video he wrote this:

Code:

  struct
  {
  int member : 3;
  }d1;
  d1.member=7;

and he said that: d1.member is can stored 7(cause you specify 3bits to use it )and he run the program and the out put was 7 it works normally.
the problem when i copy his program in my compiler code blocks
i surprised because the out put was -8 and it's different from the first ...i creat some program to understand what happning and finally i found that out :
when we write this

Code:

  struct 
  {
   int member : 3;
  //it means we use just 2 bits for store the number and last one is     for sign(negative or positive number).
   }d1;

so the greatest value we can store it in d1.member is 3 not 7.

where's the right ?

The signedness of bit fields is implementation-defined. That is, a compiler could treat it as either signed or unsigned--yours treats it as signed whereas the compiler that the teacher in the video used treats it as unsigned. Either way is acceptable according to the standards.

Use an explicit "signed" or "unsigned" type modifier if the signedness matters to you.