C++ - Bit shifting returning negative numbers

Question

I'm having some trouble trying to use bit shifting in C++.

I have a char a = 160, and char b = 0. The value of a is 0b10100000. I want to shift the leftmost bits of a into b, so that b would have the value 2 (0b10). By my logic, this should do it:

b = (b << 2) | (a >> 6);

However, when I check the value in b, it's -2 and not 2, as I would expect it to be. Casting it to an unsigned char changes it's value to 254.

What am I doing wrong?

`(b << 2)` is totally unnecessary if `b==0`; what happens if you just `char b= a >> 6;`? — Marcus Müller, Oct 22 '15 at 09:12
@JoachimPileborg Was thinking exactly the same, but I don't know any machine where integers would have a zero highest bit describe a negative number — Marcus Müller, Oct 22 '15 at 09:13
Probably your `char a = 160` is not really 160, but something like -96 (hint: if char are signed the maximum value is 127, and 160 is past this limit). — 6502, Oct 22 '15 at 09:14
@JoachimPileborg ah nevermind, I'm stupid; 160 of course has the highest bit set to one.... — Marcus Müller, Oct 22 '15 at 09:14

score 1 · Accepted Answer · answered Oct 22 '15 at 09:15

1

Both types must be unsigned:

unsigned char a = 160;
unsigned char b = 0;
b = (b << 2) | (a >> 6);

is exactly the same as:

unsigned char c = a >> 6;

answered Oct 22 '15 at 09:15

Hubi

1,629
15
20

C++ - Bit shifting returning negative numbers

1 Answers1