I wasn't trying to start anything, YES, I should have AND'ed the mask to isolate the bit(s) of interest before XORing them. But seriously at a fundamental level such as bit manipulation, much of the logic is all the same no matter what language you are using.

BTW) Embedded C does allow for inline assembly capabilities which would allow any of those instructions to be applied.


When and Why Do We Need the AND Operation in extracting specific bits?

Example1: (Need AND to read Bit)

Suppose the data you read has a decimal value of 38 and you wanted to only look at BIT 2

The Binary value for 38 is b'00100110'

Data = b'00100110'

So to look at BIT 2, you create a Mask with that bit set to a '1'

Mask = b'00000100'

When you AND the Data with your Mask the result will be 4 or the binary value of b'00000100'

If you want to determine if BIT 2 is a 1 or a 0 you just look to see if the result is greater than 0 or equal to 0.
If it's greater than 0 then BIT 2 is a 1 otherwise if the result was equal to 0 then BIT 2 is a 0.

---------------------------------------------------------------------------------------------------------------------------------

Example2: (Don't need AND to set BIT)

Let's say you wanted BIT 3 to be a 1 in the Same Data value of 38 without changing any of the other bits.

Data = b'00100110'

You would still create a Mask for BIT 3

Mask = b'00001000'

To set BIT 3 to a 1, Simply OR the Data with your Mask the result will be 46 or the binary value of b'00101110'

---------------------------------------------------------------------------------------------------------------------------------

Example3: (Need AND to clear BIT)

Let's say you wanted BIT 5 to be a 0 in the Same Data value of 46 without changing any of the other bits.

Data = b'00101110'

create a Mask for BIT 5

Mask = b'00100000'

Invert the Mask value by XORing it with b'11111111'

Inv_Mask = Mask XOR b'11111111' = b'11011111'

To set BIT5 to a 0, Simply AND the Data with your Inv_Mask the result will be 14 or the binary value of b'00001110'

 

I am still confused, I see three types of bit extraction examples.


Extracting a Single Bit:

Let's say we have the binary number 10110110, and we want to extract the 4th bit from the right (starting from 0). The 4th bit is 1.


Result = 0000 0001


Extracting a Range of Bits:

If we want to extract a range of bits, for example, bits 3 to 6 from the right in the binary number 11011010, we would get 1011. This involves extracting bits 3, 4, 5, and 6.


Result = 0001011


Extracting Specific Bits:

To extract specific bits, such as bit 7, bit 5, and bit 3 from the binary number 11011010.

So, the extracted bits are 101.


Result = 0000 0101

A fairly generic approach would be to employ a simple function which extracts arbitrary bits from a machine-word.

unsigned long long fetch_bits(unsigned long long value, int start, int length) {
  enum { bpw = sizeof(unsigned long long) * 8 };
  if (start < 0 || length < 0 || (start + length >= bpw))
    return -1;
  return (value << (bpw - start - length)) >> (bpw - length);
}

Note: in this case masking is not necessary!