I'm trying to convert a 128-bit unsigned integer stored as an array of 4 unsigned ints to the decimal string representation in C:
unsigned int src[] = { 0x12345678, 0x90abcdef, 0xfedcba90, 0x8765421 };
printf("%s", some_func(src)); // gives "53072739890371098123344"
(The input and output examples above are completely fictional; I have no idea what that input would produce.)
If I was going to hex, binary or octal, this would be a simple matter of masks and bit shifts to peel of the least significant characters. However, it seems to me that I need to do base-10 division. Unfortunately, I can't remember how to do that across multiple ints, and the system I'm using doesn't support data types larger than 32-bits, so using a 128-bit type is not possible. Using a different language is also out, and I'd rather avoid a big number library just for this one operation.
In computer architecture, 128-bit integers, memory addresses, or other data units are those that are 128 bits (16 octets) wide. Also, 128-bit central processing unit (CPU) and arithmetic logic unit (ALU) architectures are those that are based on registers, address buses, or data buses of that size.
The 128-bit data type can handle up to 31 significant digits (compared to 17 handled by the 64-bit long double).
Solution: Use sprintf() function. You can also write your own function using ASCII values of numbers.
Division is not necessary:
#include <string.h>
#include <stdio.h>
typedef unsigned long uint32;
/* N[0] - contains least significant bits, N[3] - most significant */
char* Bin128ToDec(const uint32 N[4])
{
  // log10(x) = log2(x) / log2(10) ~= log2(x) / 3.322
  static char s[128 / 3 + 1 + 1];
  uint32 n[4];
  char* p = s;
  int i;
  memset(s, '0', sizeof(s) - 1);
  s[sizeof(s) - 1] = '\0';
  memcpy(n, N, sizeof(n));
  for (i = 0; i < 128; i++)
  {
    int j, carry;
    carry = (n[3] >= 0x80000000);
    // Shift n[] left, doubling it
    n[3] = ((n[3] << 1) & 0xFFFFFFFF) + (n[2] >= 0x80000000);
    n[2] = ((n[2] << 1) & 0xFFFFFFFF) + (n[1] >= 0x80000000);
    n[1] = ((n[1] << 1) & 0xFFFFFFFF) + (n[0] >= 0x80000000);
    n[0] = ((n[0] << 1) & 0xFFFFFFFF);
    // Add s[] to itself in decimal, doubling it
    for (j = sizeof(s) - 2; j >= 0; j--)
    {
      s[j] += s[j] - '0' + carry;
      carry = (s[j] > '9');
      if (carry)
      {
        s[j] -= 10;
      }
    }
  }
  while ((p[0] == '0') && (p < &s[sizeof(s) - 2]))
  {
    p++;
  }
  return p;
}
int main(void)
{
  static const uint32 testData[][4] =
  {
    { 0, 0, 0, 0 },
    { 1048576, 0, 0, 0 },
    { 0xFFFFFFFF, 0, 0, 0 },
    { 0, 1, 0, 0 },
    { 0x12345678, 0x90abcdef, 0xfedcba90, 0x8765421 }
  };
  printf("%s\n", Bin128ToDec(testData[0]));
  printf("%s\n", Bin128ToDec(testData[1]));
  printf("%s\n", Bin128ToDec(testData[2]));
  printf("%s\n", Bin128ToDec(testData[3]));
  printf("%s\n", Bin128ToDec(testData[4]));
  return 0;
}
Output:
0
1048576
4294967295
4294967296
11248221411398543556294285637029484152
                        If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With