Reverse Bit Operations

[chillidog23]

Hi,
I am looking for a C++ function to reverse the bits of a variable. For example, assume testvar = 11100110, then reverse(testvar) = 01100111. Does anyone know of such a function? Thanks.

 

[chillidog23]

I should've mentioned that testvar is represented as a binary above, i.e. testvar = 11100110, and reverse(testvar) = 01100111.

 

[chillidog23]

Sorry, I messed up again: testvar = 11100110b, and reverse(testvar) = 01100111b

 

[chillidog23]

I also meant C, not C++... sorry again

 

[ArkM]

Of course, no such library function but there are full set of bitwise operators in C and C++: <<, >>, &, |, ^, ~ (+ loop statements.
You may use more sofisticated approach with byte by byte substitution too.
Alas, no binary constants with b suffix in C++...

 

[lionelhill]

If you need to do this very fast for any reason, it would be worth making a 256-byte lookup table with the results already reversed. Even if you're reversing much larger numbers than 8 bits, the look-up would pay off as you can reverse the bytes of a big number very easily, and then reverse the bits in each byte using the look-up. Reversing bits by repeated shifts and tests is quite a lot of cycles per byte.

 

[cpjust]

Isn't there also an assembly instruction on some chips that reverses bits? Is so, you could use that (as long as you don't care about multi-platformedness).

 

[lionelhill]

There wasn't in the 8086 family up to the pentium. I don't know if there's something whizzy since then. There were bscan instructions, forwards and backwards, that found the position of a bit in a register, but they were incredibly, outrageously slow; basically they were a repeated 1-bit shift and test, and ran just about as efficiently.

 

[Salem]

If you want an obfuscated answer, then try

http://graphics.stanford.edu/~seander/bithacks.html

--
If you dance barefoot on the broken glass of undefined behaviour, you've got to expect the occasional cut.

 

[ArkM]

Well, look at this (portable - 16..24..32..64.. monster:

static unsigned int rb[256] = /* uint contains 8*n bits */
{
0x00,0x80,0x40,0xC0,0x20,0xA0,0x60,0xE0,0x10,0x90,0x50,0xD0,0x30,0xB0,0x70,0xF0,
0x08,0x88,0x48,0xC8,0x28,0xA8,0x68,0xE8,0x18,0x98,0x58,0xD8,0x38,0xB8,0x78,0xF8,
0x04,0x84,0x44,0xC4,0x24,0xA4,0x64,0xE4,0x14,0x94,0x54,0xD4,0x34,0xB4,0x74,0xF4,
0x0C,0x8C,0x4C,0xCC,0x2C,0xAC,0x6C,0xEC,0x1C,0x9C,0x5C,0xDC,0x3C,0xBC,0x7C,0xFC,
0x02,0x82,0x42,0xC2,0x22,0xA2,0x62,0xE2,0x12,0x92,0x52,0xD2,0x32,0xB2,0x72,0xF2,
0x0A,0x8A,0x4A,0xCA,0x2A,0xAA,0x6A,0xEA,0x1A,0x9A,0x5A,0xDA,0x3A,0xBA,0x7A,0xFA,
0x06,0x86,0x46,0xC6,0x26,0xA6,0x66,0xE6,0x16,0x96,0x56,0xD6,0x36,0xB6,0x76,0xF6,
0x0E,0x8E,0x4E,0xCE,0x2E,0xAE,0x6E,0xEE,0x1E,0x9E,0x5E,0xDE,0x3E,0xBE,0x7E,0xFE,
0x01,0x81,0x41,0xC1,0x21,0xA1,0x61,0xE1,0x11,0x91,0x51,0xD1,0x31,0xB1,0x71,0xF1,
0x09,0x89,0x49,0xC9,0x29,0xA9,0x69,0xE9,0x19,0x99,0x59,0xD9,0x39,0xB9,0x79,0xF9,
0x05,0x85,0x45,0xC5,0x25,0xA5,0x65,0xE5,0x15,0x95,0x55,0xD5,0x35,0xB5,0x75,0xF5,
0x0D,0x8D,0x4D,0xCD,0x2D,0xAD,0x6D,0xED,0x1D,0x9D,0x5D,0xDD,0x3D,0xBD,0x7D,0xFD,
0x03,0x83,0x43,0xC3,0x23,0xA3,0x63,0xE3,0x13,0x93,0x53,0xD3,0x33,0xB3,0x73,0xF3,
0x0B,0x8B,0x4B,0xCB,0x2B,0xAB,0x6B,0xEB,0x1B,0x9B,0x5B,0xDB,0x3B,0xBB,0x7B,0xFB,
0x07,0x87,0x47,0xC7,0x27,0xA7,0x67,0xE7,0x17,0x97,0x57,0xD7,0x37,0xB7,0x77,0xF7,
0x0F,0x8F,0x4F,0xCF,0x2F,0xAF,0x6F,0xEF,0x1F,0x9F,0x5F,0xDF,0x3F,0xBF,0x7F,0xFF,
};
/* this routine ignores byte structure (and addressing scheme) */
unsigned int revbits(unsigned int u)
{
  int	i, j, k;
  unsigned int a, b;
  unsigned int r = 0;

  static int n;
  static unsigned int uu = 0x100u;

  /* octets per unsigned int?
   it's interesting: can't calc with preprocessor!
  */

  if (uu) /* the 1st call... */
  for (n = 2; uu <<= 8; ++n) /* not thread-safe code! */
     ;
  k = n;

  for (i = 0, j = 8*(n-1); k > 1; k -= 2, i += 8, j -=8)
  {
    a = rb[(u >> i)&0xFFu];
    b = rb[(u >> j)&0xFFu];
    r |= (b << i);
    r |= (a << j); /* next two reverted octets swapped */
  }
  if (k)	/* (now exotic system with odd num of octets per uint;) */
     r |= (rb[(u >> i)&0xFFu] << i);
  return r;
}

 

[lionelhill]

The obfuscated answers are truly remarkable but I'd love to see their execution times compared to simple look-ups and a well-placed bswap. Multiplication! Good gracious.