/*
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 2 of the License, or
 * (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
 */

// Fixed-point math routines (signed 19.12)
// Largest positive value:  524287.999755859375
// Smallest positive value: 0.000244140625

#define FRAC_BITS      12
#define INT_MASK       0x7FFFF000 // 20 bits
#define FIXED_1        4096    // (1 << FRAC_BITS)
#define FIXED_255      1044480 // (255 << FRAC_BITS)
#define FIXED_HALF     2048    // (fixed_t)(0.5 * (float)(1L << FRAC_BITS) + 0.5)
#define FIXED_EPSILON  1

#define ROUND_FIXED_TO_INT(x) ((int)(x < 0 ? 0 : (x > FIXED_255) ? 255 : fixed_to_int(x + FIXED_HALF)))

typedef int32_t fixed_t;

static inline fixed_t int_to_fixed(int32_t x) {
  return x << FRAC_BITS;
}

static inline int32_t fixed_to_int(fixed_t x) {
  return x >> FRAC_BITS;
}

static inline fixed_t float_to_fixed(float x) {
  return ((fixed_t)((x) * (float)(1L << FRAC_BITS) + 0.5));
}

static inline float fixed_to_float(fixed_t x) {
  return ((float)((x) / (float)(1L << FRAC_BITS)));
}

#if defined(__GNUC__)
# if defined(__arm__)
static inline fixed_t fixed_mul(fixed_t x, fixed_t y) {
  fixed_t __hi, __lo, __result;
  __asm__ __volatile__(
    "smull %0, %1, %3, %4\n\t"
    "movs %0, %0, lsr %5\n\t"
    "adc %2, %0, %1, lsl %6"
    : "=&r" (__lo), "=&r" (__hi), "=r" (__result)
    : "%r" (x), "r" (y), "M" (FRAC_BITS), "M" (32 - (FRAC_BITS))
    : "cc"
  );
  return __result;
}
# elif defined(__i386__) || defined(__x86_64__)
// This improves fixed-point performance about 15-20% on x86
static inline fixed_t fixed_mul(fixed_t x, fixed_t y)
{
  fixed_t __hi, __lo;
  __asm__ __volatile__(
    "imull %3\n"
    "shrdl %4, %1, %0"
    : "=a"(__lo), "=d"(__hi)
    : "%a"(x), "rm"(y), "I"(FRAC_BITS)
    : "cc"
  );
  return __lo;
}
# elif defined(PADRE) // Sparc ReadyNAS
static inline fixed_t fixed_mul(fixed_t x, fixed_t y)
{
  fixed_t __hi, __lo, __result;
  __asm__ __volatile__(
    " nop\n"
    " nop\n"
    " smul %3, %4, %0\n"
    " mov %%y, %1\n"
    " srl %0, %5, %0\n"
    " sll %1, %6, %1\n"
    " add %0, %1, %2\n"
    : "=&r" (__lo), "=&r" (__hi), "=r" (__result)
    : "%r" (x), "r" (y), "M" (FRAC_BITS), "M" (32 - (FRAC_BITS))
    : "cc"
  );
  return __result;
}
# else // Other gcc platform
static inline fixed_t fixed_mul(fixed_t x, fixed_t y) {
  return (fixed_t)(((int64_t)x * y) >> FRAC_BITS);
}
# endif
#elif defined(_MSC_VER) // x86 Windows
static inline fixed_t fixed_mul(fixed_t x, fixed_t y) {
  enum {
    fracbits = FRAC_BITS
  };
  
  __asm {
    mov eax, x
    imul y
    shrd eax, edx, fracbits
  }
  // eax is returned automatically
}
#else // Other non-gcc platform
static inline fixed_t fixed_mul(fixed_t x, fixed_t y) {
  return (fixed_t)(((int64_t)x * y) >> FRAC_BITS);
}
#endif

// XXX ARM version from http://me.henri.net/fp-div.html ?
static inline fixed_t fixed_div(fixed_t x, fixed_t y) {
  return (fixed_t)(((int64_t)x << FRAC_BITS) / y);
}

static inline fixed_t fixed_floor(fixed_t x) {
  return x & INT_MASK;
}