bitwise macro for getting a bit not working

Question

I have a short variable (16 bits) and an index (unsigned char). I need a macro that returns the indexth in my variable data. This is what I got:

#define GETBIT(data, index) data & 1 << index

And how I use it:

unsigned char i;
short * twobytes = (short *) calloc(1, sizeof(short));
twobytes =  ((char * )buffer + *currentIndex);
while (codeLength != 0)
{
    i = GETBIT(code, codeLength--);
    *twobytes = SETBIT(*twobytes, *currentBitIndex, i);
    (*currentBitIndex)++;
    if (*currentBitIndex == 8) {
        (*currentIndex)++;
        (*currentBitIndex) %= 8;
    }
}

For some reason i always equals to 0 in my test cases, where it sometimes should equal 1.

What am I doing wrong and how should I fix it?

Thanks.

In the first case `00` and in the second one `01`, in both cases I get just zeroes. — shoham, Jul 15 '14 at 17:43
@dari: `<<` has higher precedence than `&`. shoham: `01 & 1 << 1` is `01 & 2` which is `0` and `codeLength` is never `0` when you use `GETBIT`. — mafso, Jul 15 '14 at 17:45
Also, always put parenthesis around macros (if they evaluate to a value) and around the arguments: `#define GETBIT(data, index) ((data) & 1<<(index))`. — mafso, Jul 15 '14 at 17:48

score 0 · Answer 1 · answered Jul 15 '14 at 17:49

0

Nevermind, thanks, instead of codeLength-- I should've done --codeLength.

answered Jul 15 '14 at 17:49

shoham

792
2
12
30

score 0 · Answer 2 · answered Jul 15 '14 at 17:49

0

Example:

$ cat qq.c 
#include <stdio.h>

#define GETBIT(data, index) ((data & (1 << index)) == 0 ? 0 : 1)

int main() {
    const int x = 14;
    int i;
    for (i = 0; i < 32; i++) {
        printf("%d ", GETBIT(x, i));
    }
    printf("\n");
    return 0;
}

Run:

$ gcc -Wall qq.c && ./a.out 
0 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

answered Jul 15 '14 at 17:49

paq

180
6

`#define GETBIT(data, index) (((uintmax_t)(data) >> (index)) & 1)` is a little more general and a little simpler. If you are sure that the `data` is never any larger than `int`, then `unsigned` instead of `uintmax_t`. It's generally worth avoiding shifts of signed values ! – Jul 16 '14 at 00:07
http://stackoverflow.com/a/4009954/2557000 (excerpt 6.5.7/4). But, of course, bitwise operations require to be very careful :) – paq Jul 16 '14 at 10:16

bitwise macro for getting a bit not working

2 Answers2